Remerciements

Cette thèse a été pour moi une période très enrichissante tant scientifiquement que humainement. Je souhaite remercier toutes les personnes qui y ont contribué de près ou de loin, et qui ont permis de rendre ces trois années aussi riches que ce qu’elles ont été.

Je souhaite tout d’abord remercier le Dr. Azzedine Bousseksou de m’avoir accueillie au sein de son établissement.

Je remercie également tous les membres de mon jury pour avoir évalué mon travail : Dr. Catherine Belle et Dr. Yves Le Mest, merci d’avoir accepté d’être rapporteurs de ma thèse, Dr. Eric Benoist, Dr. Katell Sénéchal-David, et Dr. Stéphane Torelli, merci d’avoir accepté d’être examinateurs de ma thèse.

Je souhaite maintenant remercier la personne sans qui tout cela n’aurait pas été possible : Christelle. Merci pour ta confiance, merci pour ta patience, merci d’avoir cru en moi !!! Tu m’as transmis tellement de choses… Tu es un puits de connaissances et d’idées inépuisable ! Tu m’as permis de faire ce que j’avais envie, de tester, d’expérimenter, etc. même si parfois j’aurais pu m’abstenir… Tu m’as également laissé faire des enseignements, tu m’as fait découvrir le monde des congrès, j’ai pu avoir des stagiaires, j’ai pu aller en Argentine en collaboration, etc. (la liste est tellement longue que je vais m’arrêter là !) Je ne te remercierai jamais assez pour tout ça ! Cette thèse a été pour moi un super moment de ma vie (j’espère que pour toi, ça n’a pas été un trop long calvaire ;) ) Il ne me reste plus qu’une chose à espérer…que tu tiennes ta promesse le 1er décembre…

Peter, un grand merci à toi d’être venu me proposer de joindre ton équipe. Mais surtout, je te remercie pour toutes ces discussions enrichissantes, ces conseils toujours très avisés, ta bonne humeur quotidienne, ton savoir, … A nouveau, la langue française n’est pas suffisamment riche pour te remercier de tout ce que je voudrais… Alors tout simplement, merci pour tout !

Je tiens également à remercier tous les collaborateurs qui m’ont gentiment donné leur ligand ! Pascale Delangle, Raphaël Tripier, Laurent Lisnard, Clotilde Policar, Claude Gros et Olga Iranzo, un grand merci, car sans vous, je serais encore (après ces 3 ans de thèse) au stade de la synthèse du premier ligand !! Merci aussi à Sabrina Noël pour avoir synthétiser suffisamment de L2 pour qu’il en reste pour moi !!!

Un grand merci à toutes les personnes des services scientifiques, sans qui, les manips auraient été plus que difficiles … Tout d’abord, un immense merci à Lionel Rechignat pour m’avoir TRES patiemment expliqué la RPE, malgré mes milliards de questions tordues… pour avoir passé mes interminables séries de tubes et aussi pour avoir toujours été derrière moi, même une fois que j’ai presque pris mon indépendance en RPE ! Isabelle Kieffer, Olivier Proux et Denis Testemale, les sessions à l’ESRF sur FAME ont toujours été des moments plus qu’attendus ! Vous m’avez expliqué tellement de concepts scientifiques (mais pas que !!), vous avez toujours pris le temps pour moi, pour que je comprenne, pour nous faire visiter les lignes, pour nous montrer aussi la construction de la nouvelle ligne… sans parler des sessions bonbons, des sessions musiques pourries et quizz musical, des appels téléphoniques très étranges, etc… Merci pour tout à vous trois, et j’espère revenir sur FAME !!  Merci à Emmanuel Guillon et Stéphanie Sayen pour les moments passés à l’ESRF, les explications sur le XAS et la « ballade course » pour aller voir le synchrotron d’en haut !

~ 1 ~

Merci aussi à Vanessa Soldan pour ton aide précieuse en TEM. Merci également à Christian Bijani pour ton aide en RMN. Enfin, je tiens à remercier chaleureusement le service informatique du LCC, Jérôme et Fabrice, pour avoir toujours su me sortir des « boulettes » et autres problèmes informatiques (la plupart du temps occasionnés par moi-même !!), ainsi que le service administratif, Béatrice et Florence, pour votre aide si précieuse et votre gentillesse.

Je souhaite maintenant remercier toutes les personnes de l’équipe F. Manu, merci à toi pour ta bonne humeur, et la bonne ambiance que tu mets au sein de cette équipe !! Ça a été un réel plaisir de te connaitre. Fabrice, merci pour ta joie de vivre, pour les moments (comment les appeler ???) « délires » ou « pétage de plomb » à l’ESRF. On y a vraiment passé de supers séjours !!! James Bond, kidnapper de Petit Poney demandant une rançon restera quand même dans mes souvenirs !!!  Béa, merci à toi pour tes conseils, ta gentillesse et tes bonnes adresses ! Viviane, je te remercie pour ton oreille toujours attentive, ta douceur et ta gentillesse. Je te remercie également pour tout le temps que tu m’as consacré : les heures incalculables à synthétiser mes ligands pourris, mais aussi et surtout le fameux mois où j’ai essayé d’être une chimiste organicienne !!! Même avec toute ta bonne volonté, et ton enthousiasme, je n’arriverai pas à faire de la « vraie chimie » Viviane, et je pense que c’est plus sécuritaire pour tout le monde !! ;) Merci aussi à toi Laurent, pour avoir remis en bon fonctionnement tout le labo !!! Ton aide a été très précieuse et m’a fait gagner un temps fou !! Merci aussi pour les découvertes de films « cultes » à l’ESRF, et je n’oublierai pas que tu « adores le pognon » !!! Je souhaite maintenant remercier tous les étudiants/post docs de l’équipe, en commençant par Adam. Tu m’as très gentiment et rapidement intégrée au sein de l’équipe et je t’en remercie. Ta maladresse légendaire restera gravée dans ma mémoire !! Marie, merci à toi d’avoir toujours su m’écouter et me conseiller. Clémence, Valentina et Elena, trop de choses à dire alors voir ci-dessous !! Omar, merci pour ta patience avec mon anglais (oui oui, Jacques Chirac restera ton meilleur ami, j’ai tout compris !!) et surtout merci pour tous ces bons moments passés ensemble, sans oublier tes « supers » conseils (les bombes par exemple !! ;) ). Alex, merci à toi pour toutes ces conversations de pause-café plus tordues les unes que les autres, mais tellement drôles !! Sara, merci pour ta gentillesse et tes conseils et explications sur la biologie !! Olena, merci également à toi pour ta présence, pour nos échanges culturels, pour toutes ces découvertes culinaires, … Merci aussi à Sylvain et Valerii (pas de l’équipe F, mais on vous accepte quand même !!). Merci également à Olivia, Daniel, Melisa, Mireia, Carine, Inga et Simon, et toutes les personnes qui sont venues participer au bon fonctionnement et à la bonne ambiance de cette équipe !! Je tiens aussi à remercier « mes » 3 stagiaires qui m’ont fait confiance (oui vous n’aviez pas le choix, mais quand même !!) : Claudia, Tiriana et Agnès, ça a été un réel plaisir de travailler avec vous, et je ne vous oublierai pas. Merci du fond du cœur à toute l’équipe F, les week ends d’équipe, les Noël, etc., J’ai passé presque 4 superbes années avec vous !

Je voudrais maintenant remercier toute l’équipe pédagogique et notamment Jean-Luc et Barbora, pour m’avoir permis d’être presque une prof pendant 2 ans !! J’ai beaucoup appris, et c’est une super expérience que vous m’avez fait vivre. Merci pour votre confiance.

Ahora, quisiera agradecer a las personas de Argentina. Sandra, gracias por tu acogida tan cálida en tu equipo. Gracias también por tu gentileza y tus explicaciones sobre todo!! Claudia, fuiste quasi como una mama argentina para mí. Me hiciste descubrir muchas cosas científicas, pero también cosas argentinas y humanas. Siempre eres muy linda conmigo, y te lo agradezco. Espero que podamos vernos pronto, en Rosario, o aquí, en Francia. Nunca te olvidare… Gracias también a todo el equipo que fue muy lindo conmigo. Un gracias especial para mis amigos colombianos de Argentina!! Angel, mi compañera de habitación, tantas cosas por decir pero no hay bastante palabras… Nos veremos

~ 2 ~ pronto, y espero que podrás leer más que la primera página de este libro !!! Gracias por todo… Felipe, Pilar, Juan, Angelito, fueron tan cariñosos conmigo… Esas tres semanas quedaran por siempre en mi corazón !! Les extraño un montón…

Le club jeune de la SCF en Midi-Pyrénées a été aussi un bel investissement au cours de ma thèse. J’ai fait de belles rencontres, appris beaucoup… Merci particulièrement à toi Claudia, à toi Morgane et à toi Jérémy !! J’ai beaucoup aimé travailler avec vous !! Lydie, je te remercie également pour tout ce que tu as fait pour le club de jeunes, mais surtout pour moi !! J’espère que nos routes se recroiseront à Chimie et Terroir ou ailleurs !

Maintenant, je veux remercier mes amis, toujours présents pour moi, en toutes circonstances ! Lilia et Muriel, un immense merci à toutes les deux. Vous m’avez toujours motivée, toujours écoutée dans les bons mais aussi les moins bons moments… Toujours présentes, bien qu’à l’autre bout de la France (voire du monde par moments), cette thèse n’aurait jamais pu aussi bien se passer sans vous… Merci du fond du cœur pour tout les filles !! Bien sûr, Antoine, Louis (alias Titou et Loulou) merci à vous aussi pour tous ces bons moments passés ensemble !!! Vivement les prochains.

Clémence, à mon tour de te remercier… Et comme tu me l’as si bien dit, les mots sont ternes à côté de ce que mon cœur voudrait dire… (et oui, même avec tous mes mots du sud-ouest, je n’y arrive pas !!!) On a formé une chouette équipe… à reformer en collaboration à l’ESRF ou ailleurs !! ;) Grace à toi, ma thèse a été une super expérience !! Beaucoup de travail, mais aussi beaucoup de rigolades, de moments inoubliables (hein ma Clem Blonde !!), etc. etc. Tu as toujours été de bons conseils, toujours de bonne humeur, toujours là, tout simplement…Bref, un immense merci du fond du cœur pour tout !!

Valentina, ma coinquilina, grazie mille !! Je suis réellement heureuse que nos chemins se soient croisés (j’en deviendrai presque poète !!). Tellement de bons moments passés ensemble (et comme tu dis : la vie quoi !!)… et nos to do lists interminables de choses à faire après la journée…  nos sessions ralages vont aussi me manquer je crois… Et oui, bien que tu ne sois pas d’accord avec ça, tu es devenue une vraie française !!! Un gros bisou à la française pour te remercier !! Et surtout, prends soin d’Ounette et de mon bureau !! ;)

Elena, muchas gracias por todo !! Toujours de bonne humeur, toujours à voir le bon côté des choses … Tu es l’oreille attentive de l’équipe et je ne t’en remercierai jamais assez… Tu m’as également apporté une grande aide, merci beaucoup !!! J’ai passé un super moment au LCC (et à Toulouse ou ailleurs!!), et c’est en partie grâce à toi ! Merci d’être toi, et d’être toujours là pour moi…

Julie, à mon tour d’évoquer les fameux zumbapéros… C’est quand même une superbe invention !!! Sans oublier les parties de Just Dance ! On aura bien rigolé ! Toi aussi, tu as toujours été présente pour moi. Merci pour tout et à très vite j’espère…

Emilie ma tikou, et Eloise, merci les filles de m’avoir accompagnée tout au long de ces trois/quatre années (et de ma vie, mais ce n’est pas le sujet !!! ;) ) Vivement de vous revoir !!

Maintenant, je souhaite remercier mes parents chéris, sans qui, rien n’aurait pu être possible. Vous m’avez toujours poussée pour que je réalise mes rêves, pour que je donne le meilleur… Vous êtes des parents en or, toujours là pour moi. Et j’espère qu’avec tout ce que j’ai pu vous saouler, le peptide amyloide β n’a plus de secret pour vous !!! Un immense merci pour tout… Je vous aime…

Ahhhhh Emilie et Eva, mes frangines d’amour !! Si quand on était petites, on nous avait dit qu’on serait incapables de vivre les unes sans les autres, on aurait bien ri !!! Mais qu’est-ce qu’on aurait rrrrri ! Et pourtant… Je ne serai jamais arrivée à faire tout ça sans vous ! Vous m’êtes

~ 3 ~ indispensables !! Merci d’être toujours là, pour tout et pour rien, sérieusement ou en déconnant, mais toujours là !!! Rémi et Adrien, mes beaux-frères préférés, je ne vous oublie pas non plus ! Qu’est-ce qu’ils font du bien ces moments passés tous ensemble ! Vivement les prochains… Merci à vous quatre du fond du cœur… Je vous zaimeuhh…

Je ne vous oublie pas non plus, Papi Max, Mamie Solange et Mamie Jeannette ! Toujours là, toujours le sourire, toujours une oreille attentive pour une session papotage. C’est si bon d’être aussi bien entourée ! Merci également à papi Maurice et Dany et merci aussi pour les supers vacances détente à Marseillan qui m’ont permis de me ressourcer !

Un grand merci également à Dédé, Annie et Denis pour les bons moments passés ensemble !

Et comme on dit, le meilleur pour la fin…Laurent ! Tout d’abord, je voudrais saluer ton courage et ta patience sans limite (enfin, presque !! ;) ), parce que oui, je me rends bien compte que je n’ai pas toujours été facile avec toi… Tu as toujours su me rassurer, me réconforter, me changer les idées quand ce n’était pas très facile. Et bonus, tu connais toutes mes présentations par cœur (hihihi merci public !!). Tu es ma source d’inspiration au quotidien, la force qui me fait avancer. Sans toi, tout tournerait à l’envers, rien ne serait possible. Merci de faire de moi celle que je suis. Tout simplement, merci… Je t’aime tellement mon cœur !

~ 4 ~

List of abbreviations

7-OH-CCA: 7-hydroxycoumarin carboxylic acid

ACh: acethylcholine AChE: acetylcholine esterase AFM: atomic force microscopy APP: amyloid precursor protein Asc: ascorbate Asp: aspartate Aβ: Amyloid-β

B: β-alanine BBB: blood brain barrier BHE : barrière hémato-encéphalique

CCA: coumarin carboxylic acid CQ: clioquinol CSF: cerebrospinal fluid

Fz: ferrozine

EPR: electronic paramagnetic resonance Equiv.: equivalent EXAFS: extended X-ray absorption fine structure

Glu: glutamic acid

HEPES: 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid His: histidine H6A: mutation of the amino acid H, in the position 6, to the amino acid A

M: molar MPAC: metal protein attenuating compound

NMDA: N-methyl-D-aspartate NMR: nuclear magnetic resonance

PiB: Pittsburgh compound B PET: positron emission tomography

ROS: reactive oxygen species RPE : résonance paramagnétique électronique

SOD: superoxide dismutase

TEM: transmission electron microscopy ThT: Thioflavin T

XANES: X-ray absorption near edge structure XAS: X-ray absorption spectroscopy

~ 5 ~

Table of contents

Remerciements ...... 1 List of abbreviations ...... 5 Table of contents ...... 6

Résumé en français ...... 9 I- Contexte général ...... 9 I-A. La maladie d’Alzheimer ...... 9 I-B. Le peptide Aβ et les ions métalliques ...... 10 I-A.i Coordination et constantes d’affinité ...... 10 I-A.ii Production d’ERO ...... 12 I-A.iii L’agrégation du peptide Aβ ...... 13 I-C. Objectifs de la thèse ...... 14 II. La chélation du Cu ...... 15 II-A. Etat de l’art sur les ligands du Cu(II) contre la MA ...... 15 II-B. Impact de la cinétique de chélation ...... 19 II-C. Choix de la cible : Cu(I) et/ou Cu(II) ? ...... 22 III. Impact des ions Zn sur la chélation des ions Cu ...... 24 III-A. Etat de l’art sur les interactions mutuelles entre Cu, Zn et Aβ ...... 24 III-B. Effet thermodynamique de la présence des ions Zn ...... 27 III-C. Le concept du «pull-push» ...... 30 IV. Conclusion ...... 34

General introduction ...... 36

Chapter I: Context of the project ...... 39 I-A Alzheimer’s disease ...... 39 I-A.i Prevalence and symptoms ...... 39 I-A.ii Risk factors ...... 39 I-A.iii Histopathological hallmarks ...... 41 I-A.iv Diagnostic tools ...... 42 I-B The Amyloid-β peptide and metal ions ...... 45 I-B.i Aβ peptide ...... 45 I-B.ii Metal ions ...... 46

~ 6 ~

I-B.iii Interaction between Aβ and the metal ions Cu(I/II) and Zn(II) ...... 47 I-C Current treatments and therapeutic approaches ...... 52 I-D Objectives of the study ...... 54

Chapter II: Cu ion chelation ...... 63 II-A Cu ion chelators: State of art ...... 63 II-A.i The different categories of ligands ...... 65 II-B The kinetic aspect of the Cu(II) removal...... 73 II-B.i Draft of the publication ...... 73 II-B.ii Supplementary information ...... 80 II-C A Cu(I) and Cu(II) chelator ...... 94 II-C.i Summary ...... 94 II-C.ii Article ...... 97 II-C.iii Supporting information ...... 108 II-D Conclusion ...... 117

Chapter III. Impact of Zn(II) on the Cu(II) chelation ...... 126 State of the art: the mutual interactions of Cu and Zn ions on the Aβ peptide ...... 126 III-A.i Summary ...... 126 III-A.ii Perspective ...... 131 The thermodynamic study ...... 142 III-B.i Summary ...... 142 III-B.ii Article ...... 145 III-B.iii Supporting information ...... 153 The Cu(II) “pull-push” effect ...... 166 III-C.i Theoretical concept ...... 166 III-C.ii Experimental section ...... 168 III-C.iii Illustration of the “pull-push” concept ...... 170 III-C.iv Conclusion ...... 174 Conclusion ...... 176

General conclusion ...... 178

Annexes ...... 182 A- Determination of the affinity constant of Aβ peptide for Cu(II) ...... 182

~ 7 ~

A-i. Summary ...... 182 A-ii. Article ...... 185 A-iii. Supporting information ...... 193 B- Zn(II) coordination to Aβ peptide ...... 199 B-i. Summary ...... 199 B-ii. Article ...... 202 B-iii. Supporting information ...... 213 C- The first Cu(I) chelator against AD ...... 227 C-i. Summary ...... 227 C-ii. Article ...... 230 C-iii. Supporting information ...... 235

~ 8 ~

Résumé en français

Résumé en français

I- Contexte général

I-A. La maladie d’Alzheimer

La maladie d’Alzheimer (MA) est une maladie neurodégénérative découverte en 1906 par Alois Alzheimer. Elle touche plus de 35 millions de personnes à travers le monde et est la première forme de démence. Les symptômes de la MA sont relativement bien connus par le public : perte de mémoire pouvant devenir très importants, perte de la notion d’espace et de temps, parfois agressivité, etc. Cette maladie difficile pour le patient l’est également pour son entourage. Malheureusement, il n’existe à ce jour aucun traitement curatif. Seuls des traitements symptomatiques sont disponibles, visant à améliorer la vie des patients et à ralentir la progression de la maladie.

La MA est caractérisée notamment par la présence d’enchevêtrements neurofibrillaires intra- neuronaux de protéine Tau hyperphosphorylée (Figure I-1). La protéine Tau, nécessaire à l’assemblage et à la stabilisation des microtubules, est hyperphosphorylée dans le cas de la MA. Elle s’agrège et s’accumule en enchevêtrements neurofibrillaires. Elle n’est donc plus apte à remplir son rôle, entraînant la mort des neurones. Un autre marqueur important est la présence de plaques séniles ou plaques amyloïdes dans le cerveau (Figure I-1). Ces plaques se situent dans les fentes synaptiques, notamment dans l’hippocampe et le cortex, empêchant les connexions neuronales. Ma thèse s’intéresse aux peptides Aβ, principaux composants de ces plaques séniles ; il n’y a pas d’études concernant la protéine Tau.

Figure I-1. Schéma représentant les deux évènements importants de la MA : les enchevêtrements neurofibrillaires ainsi que les plaques séniles.

~ 9 ~

Résumé en français

I-B. Le peptide Aβ et les ions métalliques

Les plaques séniles sont majoritairement constituées de peptide Amyloïde-β (Aβ). Le peptide Aβ provient du double clivage de la protéine précurseur de l’amyloïde (APP pour Amyloid Precursor Protein) par deux enzymes, la β et la γ-sécrétase et est excrété extra-cellulairement. Aβ est un peptide d’une quarantaine d’acides aminés. Les seize premiers forment la partie hydrophile du peptide, capable de coordiner des ions métalliques. La partie C-terminale du peptide est hydrophobe et impliquée dans l’agrégation du peptide en plaques séniles. Des ions métalliques tels que les ions Cu et les ions Zn sont également présents dans ces plaques. Il y aurait 1 mM d’ions Zn et 400 µM d’ions Cu dans les plaques. Une dyshoméostasie en ions métalliques tels que les ions Cu et les ions Zn est aussi reportée dans les cerveaux atteints de la MA, les ions Cu étant déficients intra-cellulairement et en excès extra-cellulairement. Concernant les ions Zn, la tendance n’est pas claire : des concentrations supérieures mais aussi inférieures à celles dans les cerveaux sains sont détectées. Il est donc biologiquement pertinent d’étudier les interactions entre le peptide Aβ et ces ions métalliques.

I-A.i Coordination et constantes d’affinité

La coordination de ces ions métalliques avec le peptide Aβ monomérique a été très étudiée, ainsi que les constantes d’affinité associées. Les résultats présentés ici ont été effectués avec le peptide Aβ16, i.e. le peptide modèle de la séquence impliquée dans la coordination, et ce pour des raisons de solubilités restreintes du peptide entier. Concernant le Cu(II), différentes techniques spectroscopiques ont été utilisées. La coordination du Cu(II) au peptide dépend du pH. Les modes de coordination des composés 1 et 2 présentés en Figure I-2 sont ceux autour de pH 7,4 ; avec un pKa entre les deux composés à 7.7. Le composé I coordine le Cu(II) dans une géométrie plan carré : l’amine terminale, l’atome O de la liaison peptidique entre l’Asp1 et l’Ala2, l’His6 et une autre His (His13 ou His14 en équilibre). Un carboxylate venant de l’Asp1, du Glu3, de l’Asp7 ou encore du Glu11 vient compléter la sphère de coordination en position apicale. Le composé 1 est en équilibre avec le composé 2 à pH 7.4. Le composé 2 a également une géométrie plan carré autour du Cu(II). L’amine terminale est toujours impliquée dans la coordination. Ensuite, l’atome N de la liaison peptidique entre l’Asp1 et l’Ala2, l’atome O de la liaison peptidique entre l’Ala2 et le Glu3 ainsi qu’une des trois His coordine Cu(II). Il y a également un équilibre dynamique entre les trois His. Un carboxylate provenant des mêmes chaines latérales que pour le composé I vient à nouveau compléter la sphère de coordination en position apicale. Concernant la constante d’affinité d’Aβ pour Cu(II), beaucoup de valeurs très différentes ont été proposées, mais récemment un consensus a été obtenu sur la valeur de 1010 M-1 à pH 7.4. Au cours de cette thèse, une nouvelle méthode de détermination de la constante d’affinité du Cu(II) pour l’Aβ

~ 10 ~

Résumé en français et des peptides mutants par UV-vis est proposée, en collaboration avec Dr Laurent Lisnard (Université Pierre et Marie Curie, à Paris). Cette méthode est une étude par compétition, permettant un accès relativement aisé à la valeur de la constante d’affinité. Une fois que la constante d’affinité du compétiteur a été déterminée, la constante d’affinité du peptide peut être mesurée par compétition entre l’Aβ et le compétiteur. Une valeur de 1.6 109 M-1 à pH 7.1 a été déterminée. Cette valeur est en accord avec le consensus proposé de 1010 M-1 à pH 7.4 mais également avec la valeur obtenue par potentiométrie par un autre groupe. Notons que les études réalisées au cours de cette thèse sont à pH 7.1 pour des raisons de constantes d’affinité (Cu(II) et Zn(II)) de certains ligands données seulement à pH 7.1. Ensuite, différents mutants d’Aβ16 ont été étudiés par la même méthode. L’idée est ici de vérifier le mode de coordination du Cu(II) par l’Aβ par cette méthode : si un mutant a une constante d’affinité égale à celle d’Aβ, alors l’acide aminé qui a été muté n’est pas impliqué dans la coordination, alors que si la valeur de l’affinité diminue, alors l’acide aminé avant mutation était impliqué dans la coordination. Les résultats obtenus sont en accord avec les coordinations proposées en Figure I-2. Cette étude a permis de confirmer la valeur de la constante d’affinité d’Aβ pour Cu(II), mais surtout de proposer une méthode robuste de détermination de constante d’affinité (pour des molécules d’affinité comparable à celle du compétiteur).

Figure I-2. Figure représentant les différentes coordinations des ions métalliques avec Aβ ainsi que les constantes d’affinité correspondantes, à pH 7.4.

~ 11 ~

Résumé en français

Concernant le Cu(I), il est lié au peptide Aβ de façon linéaire entre deux résidus His (Figure I-2). A pH 7.4, la forme majoritaire est celle impliquant les His13 et His14. La constante d’affinité d’Aβ pour Cu(I) a été déterminée via une compétition avec la Ferrozine. Cependant, la constante d’affinité du Cu(I) pour la Ferrozine est controversée. Ainsi, deux valeurs sont proposées pour Aβ à pH physiologique, dépendant de la valeur choisie pour la Ferrozine : 1010.4 M-1 et 106.9 M-1.

L’ion Zn(II) est lié au peptide Aβ dans une géométrie tétraédrique. Dans la littérature, de nombreuses coordinations sont proposées, impliquant souvent l’amine terminale. Cependant, au cours de cette thèse, un nouveau mode de coordination a été proposé, n’impliquant pas l’amine terminale. Cette étude a été réalisée en mettant en parallèle les résultats obtenus par EXAFS, XANES et RMN. La première étape fut la détermination du nombre de ligands coordinant le Zn(II), ainsi que leur nature. Les résultats EXAFS et RMN ont permis de proposer un environnement tétraédrique. Ensuite, une série de mutants d’Aβ a été étudié par XANES et RMN en vue de déterminer quels acides aminés sont impliqués dans la coordination du Zn(II). En effet, si le signal obtenu pour le complexe Zn(II)-mutant est identique à celui obtenu pour Zn(II)-Aβ, alors l’acide aminé (avant mutation) n’est pas un ligand du Zn(II) ; au contraire, si le signal est modifié par rapport à celui de Zn(II)-Aβ, alors l’acide aminé avant mutation est impliqué dans la coordination du Zn(II). La Figure I-2 illustre le mode de coordination à pH physiologique proposé pour Zn(II) dans cette étude : l’His6, le carboxylate du Glu11, l’His13 ou l’His14, un carboxylate venant de l’Asp1, du Glu3 ou de l’Asp7. Concernant la constante d’affinité d’Aβ pour Zn (II), la valeur proposée est de 105 M-1 à pH 7.4.

I-A.ii Production d’ERO

Comme montré précédemment, le peptide Aβ peut coordiner le Zn(II), mais également le Cu dans ses deux degrés redox. Ainsi, le complexe Cu-Aβ est capable de catalyser la production d’espèces réactives de l’oxygène (ERO). Les ERO sont les produits de la réduction incomplète de l’oxygène par un

• - réducteur, généralement l’ascorbate (Figure I-3). L’anion superoxyde (O2 ) est formé, puis le peroxyde

• d’hydrogène (H2O2) et ensuite le radical hydroxyle (HO ). Ces ERO, lorsqu’elles ne sont pas régulées peuvent être très délétères pour les molécules environnantes, y compris celles constituant les parois neuronales. Dans les cerveaux atteints de la MA, les concentrations en ERO ne sont plus régulées en particulier à cause d’une surproduction catalytique d’ERO.

~ 12 ~

Résumé en français

Figure I-3. Schéma représentant la production d’ERO par le complexe Cu-Aβ ainsi que les différentes techniques utilisées pour leur étude

Il existe différentes techniques spectroscopiques pour étudier la production d’ERO. Au cours de cette thèse, deux d’entre elles ont été utilisées. Le suivi de la consommation d’ascorbate par UV-vis à 265 nm (ε = 14 500 cm-1.M-1) est corrélé à la production d’ERO. En effet, si la concentration d’ascorbate diminue, alors l’ascorbate réagit avec le complexe Cu-Aβ et déclenche la production d’ERO. Au contraire, si cette concentration est constante, alors l’ascorbate ne réagit plus/pas et donc il n’y a pas de production d’ERO. La seconde méthode utilisée ici est une étude par fluorescence détectant les ions HO•. L’acide coumarine 3-carboxylique (CCA) est ajouté au milieu réactionnel, et en présence de radicaux hydroxyles, forme le 7-OH-CCA. Le 7-OH-CCA est une molécule fluorescente. Ainsi, une détection de fluorescence est corrélée à une production de 7-OH-CCA, et donc au relargage de radicaux hydroxyles par le système de Cu. Ainsi, une production d’ERO est traduite par une détection de fluorescence.

I-A.iii L’agrégation du peptide Aβ

L’agrégation du peptide Aβ en plaques séniles est un autre évènement de la MA. Elle est décrite dans le cadre de l’hypothèse de la cascade amyloïde qui place ce phénomène au centre de la MA. Cette hypothèse est relativement bien acceptée par la communauté scientifique, bien que controversée par certains groupes. Le peptide Aβ, qui est présent sous forme monomérique dans un cerveau sain, s’agrège dans un cerveau atteint de la MA en oligomères puis en fibres, s’assemblant ensuite en plaques séniles. Les fibres sont des structures riches en feuillet β. De nombreuses études cherchent à déterminer l’impact des ions métalliques dans l’agrégation du peptide. Les résultats obtenus jusqu’à présent ne convergent pas. Avec les conditions utilisées au cours de cette thèse, les ions Cu(II)

~ 13 ~

Résumé en français stabilisent les agrégats de faibles poids moléculaires supposés être les plus toxiques, alors que les ions Zn stabilisent des espèces fibrillaires. L’étude de l’agrégation du peptide Aβ en présence ou non d’ions métalliques se fait couramment via l’intermédiaire d’un fluorophore : la thioflavine T (ThT) (Figure I-4). La ThT est une molécule constituée de deux motifs aromatiques reliés par une simple liaison. Lorsque la ThT s’intercale dans les feuillets β, ce qui empêche la libre rotation autour de la simple liaison, on observe alors une très forte exaltation de la fluorescence à 490 nm (excitation à λ = 440 ± 15 nm). Par conséquent, les suivis cinétiques de l’agrégation du peptide Aβ sont effectués par fluorescence de la ThT. Lorsque la fluorescence est détectée, il y a présence de feuillets β et donc de fibres. Ces courbes d’agrégation ont une forme sigmoïdale, corrélée aux trois étapes de l’agrégation. La première est une étape de nucléation pendant laquelle les oligomères se forment. Ensuite, une seconde étape correspond à l’élongation des oligomères et protofibres en fibres. Enfin, le plateau de la sigmoïde traduit un équilibre.

Figure I-4. Représentation de la Thioflavine T, ainsi que de la libre rotation entre les deux parties de la molécule.

I-C. Objectifs de la thèse

En résumé, le peptide Aβ agrège en oligomères, espèces désignées comme les plus toxiques des espèces présentes dans le processus d’agrégation, puis en fibres, elles-mêmes formant ensuite les plaques séniles observées dans les fentes synaptiques. Des ions métalliques tels que le Cu et le Zn interagissent avec le peptide. Les ions Cu catalysent la production d’ERO même lorsqu’ils sont liés au peptide. De plus, dans les conditions utilisées dans cette thèse, ils stabiliseraient les oligomères ou les agrégats de faible poids moléculaire, alors que les ions Zn, inertes d’un point de vue redox, semblent stabiliser des espèces fibrillaires. Ces ions Cu liés au peptide Aβ, pouvant catalyser la production d’ERO et stabilisant les agrégats de faible poids moléculaire sont par la suite appelés les « Cu toxiques ». La chélatothérapie est devenue une voie thérapeutique pour lutter contre la MA. Utiliser des ligands pour retirer les ions Cu toxiques semble prometteur et est une stratégie très étudiée. Les différents critères que doit remplir le chélateur sont : sa capacité à passer la Barrière Hémato-Encéphalique (BHE), une affinité pour les ions Cu supérieure à celle d’Aβ mais pas trop élevée pour ne pas retirer les ions Cu des

~ 14 ~

Résumé en français métalloprotéines. Dans cette thèse, différentes preuves de concept visant à mettre en lumière de nouveaux critères sont proposées. Dans une première partie, une étude bibliographique sur les ligands pour le Cu déjà existants et visant la MA est rapportée, ainsi que leur impact sur la production d’ERO et sur l’agrégation du peptide. Ensuite, une première preuve de concept concerne l’importance de la cinétique de captation du Cu(II) par le chélateur. En effet, il ne suffit pas que le chélateur ait une affinité pour le Cu supérieure à celle d’Aβ. Nous montrerons que la chélation doit être rapide. Une deuxième preuve de concept décrit une stratégie reposant sur le ciblage des deux états redox du Cu. En effet, le degré d’oxydation du Cu dans les fentes synaptiques n’étant pas connu à ce jour, cibler l’un des deux états redox pourrait s’avérer inefficace. La seconde partie de cette thèse s’attarde sur l’impact des ions Zn sur la chélation des ions Cu. En effet, puisqu’une concentration élevée en Zn(II) est retrouvée dans les plaques séniles et surtout dans la fente synaptique (respectivement 1 mM et 10 à 100 fois plus concentré que les ions Cu), il est pertinent de prendre en considération cet ion. D’abord, une étude bibliographique sur les interférences mutuelles entre les ions Zn, Cu et le peptide Aβ, en présence ou non de chélateurs, est rapportée. Leurs impacts sur la coordination, la production d’ERO et l’agrégation sont décrits. Ensuite, une preuve de concept concernant l’aspect thermodynamique de l’interaction des ions Zn dans la chélation des ions Cu est détaillée. Le chélateur doit pouvoir retirer sélectivement le Cu du peptide tout en laissant les ions Zn essentiels dans la fente synaptique. Enfin, la dernière preuve de concept propose une idée originale de chélatothérapie, le « pull-push ». Certains ligands ont une affinité pour le Cu et pour le Zn dans une gamme qui est telle que le Zn(II) devient nécessaire à la chélation des ions Cu. En absence de Zn(II), ces ligands ne sont pas efficaces dans le retrait du Cu du peptide, mais en présence de Zn(II), ils le deviennent. Le Zn(II) « tire » les ions Cu hors du peptide et les pousse dans le ligand, le Zn(II) étant alors coordiné par le peptide.

II. La chélation du Cu

II-A. Etat de l’art sur les ligands du Cu(II) contre la MA

Comme expliqué précédemment, puisque la dyshoméostasie des ions Cu est considérée comme un aspect important de la MA, l’approche de chélatothérapie et de redistribution des ions Cu contre la MA a été très étudiée. Elle consiste à utiliser des ligands capables de retirer les ions Cu du peptide ou d’interagir avec le complexe Cu-Aβ pour en changer les propriétés. Il existe à ce jour différents critères que doit respecter un ligand pour être utilisé dans le cadre de la MA. Il doit avoir une constante d’affinité pour les ions Cu supérieure à celle d’Aβ pour pouvoir les retirer du peptide. Cependant, cette

~ 15 ~

Résumé en français affinité ne doit pas être trop élevée pour ne pas retirer les ions Cu « utiles », présents dans les métalloprotéines essentielles. Ceci pourrait en effet être délétère, augmentant la dyshoméostasie déjà existante. Il est à noter que dans cette étude bibliographique, seuls les ligands capables de retirer les ions Cu d’Aβ1-x sont pris en considération. En effet, les peptides tronqués Aβ4-x et Aβ11-x commençant en positions 4 et 11, biologiquement présents dans les cerveaux de la MA, ont une constante d’affinité pour les ions Cu supérieures à celle d’Aβ1-x. Il existe deux types de ligands : les chélateurs, qui peuvent retirer le Cu du peptide et sont ensuite excrétés sous forme de complexes, ou les métallophores, qui sont des chélateurs redistribuant les ions Cu intra-cellulairement. Un autre critère important à considérer dans la conception des ligands contre la MA est que la BHE doit être perméable aux ligands, c’est-à-dire qu’ils doivent respecter au minimum les règles de Lipinski, puisque les ions Cu cibles se situent dans le cerveau. Différents paramètres doivent être respectés pour cela : le ligand ne doit pas être trop hydrophile (attention à ce qu’il ne soit pas trop hydrophobe non plus car il ne pourrait plus chélater les ions Cu), il ne doit pas avoir un poids moléculaire trop élevé, etc.

Des composés « multi-cibles » sont développés dans le cadre de la MA. Il existe deux stratégies : la partie « chélateur » peut être soit incorporée dans une seconde partie ou accrochée via un lien à cette seconde partie. Cette seconde partie peut être une partie anti-oxydante permettant de capter les ERO, peut être une partie qui cible les agrégats d’Aβ, ou encore une partie qui permet de modifier l’agrégation du peptide. On peut également trouver des parties qui permettent de passer la barrière hématoencéphalique, comme des glucoses ou des nanoparticules. Enfin, il existe aussi des « pro- chélateurs » qui sont des ligands qui ont leur partie « chélatrice » masquée. Elles sont « libérées » par une action extérieure (comme celle des sécrétases).

Il existe différentes catégories de ligands contre la MA. Il est possible de les regrouper en différentes familles, en fonction de leur structure. Une de ces familles regroupe les hydroxy/aminoquinolines, comme le clioquinol et le PBT2 (Figure II-1, en haut à gauche). Ces deux derniers sont connus pour avoir été testés en études cliniques, mais ne sont pas passés en phase III. Une raison possible de cet échec serait le manque de sélectivité pour le Cu par rapport au Zn(II). En effet, ces ligands formant des complexes 2 : 1 (ligand : ion métallique) peuvent adopter la géométrie plan carré (préférée par le Cu(II)) et tétraédrique (préférée par le Zn(II)). De nouveaux ligands basés sur le clioquinol et le PBT2 sont synthétisés, avec des modifications permettant de former des ligands 1 : 1, dont l’affinité du ligand pour le Cu(II) et la géométrie du complexe sont plus facilement contrôlables.

~ 16 ~

Résumé en français

Figure II-1. Schéma regroupant les ligands étudiés dans la bibliographie dans le cadre de la MA : les hydroxy/aminoquinolines (en haut à gauche), les ligands basés sur les benzothiazoles (en haut à droite) et les ligands homologues azotés des stilbènes (en bas).

Une autre catégorie regroupe les ligands incorporant des motifs benzothiazoles et les ligands homologues azotés des stilbènes (Figure II-1, en haut à droite et en bas, respectivement). Ces motifs peuvent s’insérer dans les feuillets-β des fibres d’Aβ. Ceci permet de cibler la chélation : si les molécules interagissent avec les fibres, alors les ions Cu ne devraient être retirés que dans les zones proches des fibres. Notons qu’avec ces ligands, des complexes 2 : 1 et 1 : 1 peuvent être formés.

~ 17 ~

Résumé en français

Figure II-2. Schéma regroupant le deuxième groupe de ligands étudiés dans la bibliographie dans le cadre de la MA : les structures de type salen, bispicen et bis(thiosemicarbazonato) (en haut à gauche), les ligands macrocycliques (en haut à droite) et les ligands peptidiques (en bas).

~ 18 ~

Résumé en français

Une troisième catégorie est la famille des ligands macrocycliques ou peptidiques (Figure II-2, en haut à droite et en bas). En effet, en général, ces ligands chélatent les ions Cu(II) avec une forte affinité, dans une géométrie relativement stable, empêchant la réduction du Cu(II) et la catalyse de production d’ERO. On forme majoritairement des complexes 1 : 1 avec ces ligands.

La dernière catégorie regroupe les ligands de type salen, bis(thiosemicarbazonato) ou encore les bispicen (Figure II-2, en haut à gauche). Ils chélatent les ions Cu(II) et forment des complexes 1 :1.

Tous ces ligands peuvent retirer les ions Cu du peptide Aβ et en général peuvent au moins stopper la production d’ERO ou induire une agrégation du type « apo » et non Cu-induite. Concernant les complexes ternaires, il est probablement plus difficile pour eux d’arrêter la production d’ERO puisque la géométrie autour des ions Cu est plus flexible, accommodant les ions Cu(II) et Cu(I) et permettant le cycle rédox Cu(II)/Cu(I). Ainsi, les complexes ternaires sont peut-être plus efficaces dans la modification de l’agrégation que dans l’arrêt de la production d’ERO. De plus, pour pouvoir concevoir un métallophore, il semble plus simple d’utiliser des complexes 1 : 1. La littérature est riche d’études de chélateurs ou métallophores contre la MA, avec des structures et actions différentes. La suite de cette thèse propose différentes preuves de concept qui constituent de nouveaux critères à prendre en compte dans la conception de ligands dans le cadre de la MA. Il faut noter que les ligands étudiés dans cette thèse ne remplissent pas, dans un premier temps, tous les critères évoqués précédemment pour des raisons de solubilité et d’études in vitro. Cependant, la sophistication de ces ligands sera à envisager dans un second temps.

II-B. Impact de la cinétique de chélation

La première preuve de concept détaillée dans cette thèse concerne l’importance de la cinétique de captation des ions Cu(II) par un ligand. Les paramètres thermodynamiques ont souvent été étudiés : le chélateur a besoin d’une constante d’affinité pour le Cu(II) supérieure à celle d’Aβ, mais pas trop élevée pour ne pas retirer les ions Cu d’autres protéines. Pour illustrer l’importance des paramètres cinétiques, deux séries de ligands macrocycliques (en collaboration avec Dr. Raphaël Tripier, Laboratoire Chimie Electrochimie Moléculaires et Chimie Analytique, Brest) basés sur les 1,4,7,10- tetraazacyclododecan (cyclen) et 1,4,8,11-tetrazacyclotetradecane (cyclam) sont étudiées (Figure II-3). Les ligands cyclen et cyclam sont connus pour leur forte affinité pour les ions Cu(II) et pour l’inertie cinétique de leur complexe.

~ 19 ~

Résumé en français

Figure II-3. Schéma des six ligands étudiés dans cette partie.

La première étude a concerné les ligands non substitués des séries, c’est-à-dire le cyclen et le cyclam. Les expériences de suivi de production d’ERO ont été réalisées et les résultats de la consommation d’ascorbate ainsi que ceux obtenus par suivi de fluorescence du 7-OH-CCA sont en accord. Si les ligands cyclen et cyclam sont ajoutés sur Cu(II)-Aβ avant l’ajout d’ascorbate, c’est-à-dire si l’on dispose d’un temps suffisamment long pour former les complexes Cu(II)-cyclen et Cu(II)-cyclam, alors il n’y a pas de production d’ERO. Cependant, si les ligands sont ajoutés en cours de production d’ERO par Cu-Aβ, alors ils ne peuvent plus retirer les ions Cu du peptide suffisamment rapidement, ne pouvant donc arrêter la production d’ERO. Il faut noter ici que la géométrie du complexe a un impact sur la cinétique de captation. En absence d’Aβ, on observe que le cyclen, mais pas le cyclam, arrive à chélater le Cu pendant la production d’ERO. Ceci est à mettre en relation avec la position du Cu(II) hors de la cavité macrocyclique dans le cas du cyclen contrairement au cyclam.

Ensuite, les ligands substitués par un ou deux bras méthylpicolinate ont été étudiés (Figure II-3). Les mêmes expériences que précédemment ont été réalisées. La Figure II-4 montre les résultats obtenus. Les panneaux A et B présentent les expériences avec temps d’incubation, c’est-à-dire avec possibilité de formation des complexes avant déclanchement de la production d’ERO. Comme pour les ligands non substitués, il n’y pas de production d’ERO, prouvant que les ligands ont retiré le Cu(II) du peptide Aβ et que les complexes formés ne forment pas d’ERO. Les panneaux C et D de la Figure II-4 présentent les expériences dans lesquelles les ligands sont ajoutés en cours de production d’ERO par le complexe Cu-Aβ. Ici, contrairement aux ligands non substitués, les ligands substitués chélatent suffisamment rapidement les ions Cu pour stopper la production d’ERO associée.

~ 20 ~

Résumé en français

Figure II-4. Cinétiques de consommation d’ascorbate suivie par UV-Vis à 265 nm avec une correction à 800 nm. Panneau A. (a) Aβ16 + Cu(II) + Asc, (b) Aβ16 + Cu(II) + cyclen + Asc, (c) Aβ16 + Cu(II) + do1pa + Asc, (d) Aβ16 + Cu(II) + do2pa + Asc. Panneau B. (a) Aβ16 + Cu(II) + Asc, (b) Aβ16 + Cu(II) + cyclam + Asc, (c) Aβ16 + Cu(II) + te1pa + Asc, (d) Aβ16 + Cu(II) + te2pa + Asc. Panneau C. (a) Asc + Aβ16 + Cu(II), (b) Asc + Aβ16 + Cu(II) + cyclen, (c) Asc + Aβ16 + Cu(II) + do1pa, (d) Asc + Aβ16 + Cu(II) + do2pa. Panneau D. (a) Asc + Aβ16 + Cu(II), (b) Asc + Aβ16 + Cu(II) + cyclam, (c) Asc + Aβ16 + Cu(II) + te1pa, (d) Asc + Aβ16 + Cu(II) + te2pa. [L] = [Aβ16] = 12 µM, [Cu(II)] = 10 µM, [Asc] = 100 µM, [HEPES] = 100 mM, pH 7.1.

Ces expériences prouvent que l’effet cinétique de captation de Cu(II) est un paramètre très important dans des approches de chélatothérapie contre la MA. En effet, bien que les ligands macrocycliques non substitués aient une très forte constante d’affinité pour les ions Cu(II), leur cinétique de captation est lente par rapport à la vitesse de production d’ERO par Cu-Aβ. Ainsi, dans le cas le plus proche de ce qui se passe in vivo, i.e. lorsque les ligands sont ajoutés en cours de production d’ERO, cyclen et cyclam ne sont pas capables de retirer Cu d’Aβ et d’arrêter la production d’ERO associée. L’ajout de bras chélateurs permet d’accélérer la captation des ions Cu(II). Ainsi, le Cu n’est plus lié à Aβ et la production d’ERO est stoppée. Cette première preuve de concept met l’accent sur l’impact de la cinétique de captation des ions Cu par des ligands sur la production d’ERO catalysée par Cu-Aβ.

~ 21 ~

Résumé en français

II-C. Choix de la cible : Cu(I) et/ou Cu(II) ?

La deuxième preuve de concept concerne le choix de la cible du chélateur. En effet, nombreux sont les ligands du Cu(II) développés pour la MA, alors que très peu de ligands ciblant le Cu(I) sont décrits dans la bibliographie. Pourtant, en théorie, hormis les difficultés pratiques à étudier les ions Cu(I), il n’y a pas d’explication à cela. En effet, à ce jour, le degré d’oxydation des ions Cu dans la fente synaptique n’est pas connu. Ainsi, comment choisir entre un ligand pour le Cu(II) et un ligand pour le Cu(I), sachant que dans les deux cas, il y a risque d’inefficacité dans la chélation ? C’est pourquoi le ligand L ciblant les deux états redox du Cu a été étudié, en collaboration avec Dr Pascale Delangle, Institut Nanosciences et Cryogénie, Grenoble. Trois résidus Histidine sont greffés sur un squelette nitrilotriacétique (Figure II-5).

Figure II-5. Schéma représentant le ligand L étudié comme chélateur du Cu(I) et du Cu(II). Un squelette nitrilotriacétique sert de support au greffage de trois résidus Histidine.

Après avoir déterminé les constantes d’affinité des complexes Cu(II)-L et Cu(I)-L, la caractérisation de ces complexes par RPE et XANES respectivement, a permis de proposer un mode de coordination pour ces complexes, en fonction du pH. A pH 7.1, le Cu(II) est coordiné en géométrie plan carré par les trois Histidines du ligand ainsi que par un amide déprotoné (Figure II-6). Le Cu(I), lui, est coordiné par les trois Histidines ainsi que par un ligand extérieur, le solvant par exemple, dans une géométrie tétraédrique (Figure II-6). Les deux modes de coordination sont bien distincts.

Figure II-6. Schéma représentant les modes de coordination du Cu(II) et du Cu(I) par le ligand L à pH 7.1.

Ensuite, une étude par voltamétrie cyclique a permis de caractériser le système. Un schéma carré ECEC (Electrochimie-Chimie-Electrochimie-Chimie) permet de le décrire. Le complexe Cu(II)-L, en

~ 22 ~

Résumé en français géométrie plan carré, est réduit en Cu(I)-L, en géométrie plan carré (E) suivi d’un réarrangement pour obtenir une géométrie tétraédrique (C). Dans l’autre sens, le complexe Cu(I)-L tétraédrique est oxydé en Cu(II)-L tétraédrique (E), suivi d’un réarrangement en Cu(II)-L plan carré (C). Il est à noter que les espèces intermédiaires, i.e. Cu(II)-L tétraédrique et Cu(I)-L plan carré ne sont pas visibles dans les conditions expérimentales utilisées. De plus, d’après les valeurs de réduction et d’oxydation de ces complexes, ils semblent être résistants à la réduction par l’ascorbate et à l’oxydation par le dioxygène.

Une fois la caractérisation des complexes effectuée, la première question a été de savoir si L pouvait retirer le Cu(II) du peptide Aβ. Pour cela, une étude par RPE a été réalisée. Il en résulte que au moins 95 % du Cu(II) est chélaté par L, laissant seulement moins de 5 % de Cu(II) sur Aβ. Ensuite, une étude par XANES a permis de déterminer que L retire également plus de 80 % de Cu(I) d’Aβ. Ces résultats sont en accord avec les valeurs de constante d’affinité mesurées. Ainsi, L est capable de retirer le Cu(I) et le Cu(II) d’Aβ. L’impact de ce chélateur sur la production d’ERO a ensuite été étudié. Le suivi de la consommation d’ascorbate a été utilisé pour cette étude, et la détection des radicaux HO• par fluorescence a permis de confirmer les résultats. Les résultats présentés ci-après sont les mêmes en présence ou non d’Aβ. Dans un premier temps, le ligand est incubé avec le Cu(II)-Aβ et une fois le complexe Cu(II)-L formé, l’ascorbate est ajouté (Figure II-7A). L’ascorbate n’est pas consommé, ainsi il n’y a pas de production d’ERO. Ensuite, le complexe Cu(I) est préparé sous Argon en présence d’ascorbate et d’Aβ (Figure II-7B). Après ouverture à l’air, il n’y a pas non plus de production d’ERO. Finalement, lorsque L est ajouté pendant la production d’ERO, c’est-à-dire pendant que les ions Cu cyclent entre l’état d’oxydation +I et +II, la production d’ERO est très fortement ralentie (Figure II-7C). Ainsi, L est capable de fortement réduire la production d’ERO que ce soit en départ Cu(II), départ Cu(I) ou un mélange des deux degrés d’oxydation.

Figure II-7. Suivi de la production d’ERO par consommation d’ascorbate, par UV-Vis à 265 nm, avec une correction de la ligne de base à 800 nm. Panneau A. Aβ + Cu(II) + Asc (courbe noire), L + Cu(II) + Asc (courbe bleue), Aβ + Cu(II) + L + Asc (courbe rouge). Panneau B. Cu(II) + Asc + Aβ + air (courbe noire), Cu(II) + Asc + L + air (courbe bleue), Cu(II) + L + Asc + Aβ + L (courbe rouge). Panneau C. Asc + Cu(II) + L (courbe bleue), Asc + Aβ + Cu(II) + L (courbe rouge). [L] = [Aβ] = 12 μM, [Cu(II)] = 10 μM, [Asc] = 100 μM, [HEPES] = 100 mM, pH 7.1. Pour les expériences du panneau B, toutes les solutions ont été déoxygénées en bullant de l’Argon et ont été ajoutées sous une légère surpression d’Argon pour garder le Cu sous son état d’oxydation +I.

~ 23 ~

Résumé en français

Cette deuxième preuve de concept décrit donc un ligand capable de chélater le Cu(II) et le Cu(I) et de produire très peu d’ERO (comparativement au peptide Aβ). Ceci semble être dû à la différence de géométrie et de coordination entre Cu(I)-L et Cu(II)-L. Ce changement est très important, empêchant le Cu de cycler et donc de produire des ERO. Cibler les deux états d’oxydation dans ce contexte est important puisque, actuellement, le degré d’oxydation du Cu dans les fentes synaptiques reste inconnu.

III. Impact des ions Zn sur la chélation des ions Cu

La première partie de ma thèse s’est focalisée sur la chélation des ions Cu dans le contexte de la MA. Après une étude de la bibliographie sur ces chélateurs, une première preuve de concept décrit l’impact de la cinétique de captation du Cu sur la production d’ERO. Ensuite, une deuxième preuve de concept détaille l’intérêt de cibler les deux états d’oxydation du Cu et propose un ligand répondant à ce critère. Bien entendu, les complexes formés ne doivent pas produire d’ERO. La seconde partie de cette thèse se concentre sur l’impact des ions Zn(II) sur la chélation du Cu(II). Ces études sont biologiquement pertinentes puisqu’il y aurait 10 à 100 fois plus de Zn(II) que d’ions Cu dans la fente synaptique.

III-A. Etat de l’art sur les interactions mutuelles entre Cu, Zn et Aβ

L’état de l’art concernant les études sur les effets des interactions mutuelles entre les ions Cu et Zn et le peptide Aβ est rapporté dans la première section de la dernière partie de la thèse. Dans les cerveaux atteints de la MA, il y a une mauvaise régulation des concentrations des ions Cu et Zn. Les ions Cu seraient trop concentrés extra-cellulairement, et déficients intra-cellulairement. Concernant les ions Zn, il est difficile de savoir si leur concentration est trop élevée ou trop faible, car ces deux résultats ont été obtenus par différents groupes. Les ions Cu et Zn ont également été retrouvés dans les plaques séniles et in vitro, les deux ions métalliques interagissent avec le peptide Aβ. Il a été montré que, séparément, ces ions pouvaient moduler l’agrégation du peptide, mais à nouveau, la tendance de cette modulation in vitro n’est pas claire. Cependant, toutes les études convergent vers l’importance du ratio métal : peptide et vers le fait que les effets des ions Cu et ceux des ions Zn sur l’agrégation sont différents. De plus, le complexe Cu(II)-Aβ catalyse la production d’ERO qui peuvent attaquer les biomolécules environnantes. C’est pour cela que la chélation des ions Cu(II) est une partie importante des recherches sur la MA depuis plusieurs années. Il existe deux grandes catégories de ligands : les

~ 24 ~

Résumé en français chélateurs et les métallophores. Les chélateurs sont des molécules capables de retirer les ions Cu du peptide ; les métallophores sont des chélateurs qui peuvent ensuite redistribuer les ions Cu intra- cellulairement. Certains de ces métallophores, comme le clioquinol et PBT2, ont été jusqu’aux tests cliniques, mais ont finalement échoué. Une des différentes hypothèses concernant ces échecs sont le manque de sélectivité des ions Cu par rapport aux ions Zn de ces ligands. En effet, ils peuvent chélater aussi bien les ions Cu que les ions Zn. Ceci met l’accent sur le fait que cibler seulement les ions Cu comme approche thérapeutique ne semble pas être suffisant et qu’il faille prendre en compte l’environnement des ions Cu. La première étape de cette approche plus réaliste est l’étude de l’impact des ions Zn(II). Premièrement, l’impact des ions Zn(II) sur la coordination des ions Cu est reporté, puis sur la production d’ERO et sur l’agrégation, et enfin sur la chélation des ions Cu.

La première partie de cet état de l’art se concentre sur les coordinations hétéro-bimétalliques. Elles ont été étudiées par différentes techniques spectroscopiques et potentiométriques. Concernant la coordination des ions Cu(II), les ions Cu(II) et Zn(II) déplacent mutuellement le site de coordination de l’autre ion (Figure III-1). Par exemple, le Zn(II) garde l’Histidine 6 comme ligand (ligand commun aux deux ions métalliques dans les complexes homo-métalliques) et le Cu(II) garde les deux autres Histidines à pH 7.1. Ainsi, le Zn(II) pousse le Cu(II) dans la coordination du composé 2 et le Cu(II) pousse le Zn(II) dans une coordination différente de celle du complexe homo-métallique, avec une seule Histidine. Concernant le Cu(I), il impose son mode de coordination, laissant une seule Histidine et non deux comme ligand pour le Zn(II) (Figure III-1). En résumé, Cu(I/II) et Zn(II) ont des interactions mutuelles sur leur site de coordination au peptide Aβ. Cependant, des études sont maintenant nécessaires pour déterminer l’impact mutuel de ces ions sur les constantes d’affinité.

~ 25 ~

Résumé en français

Figure III-1. Représentations des différents sites de coordination des ions Cu(I/II) et Zn(II) avec le peptide Aβ à pH physiologique, ainsi que des ceux des complexes hétéro-bimétalliques.

La seconde partie de cet état de l’art est concentrée sur l’impact des ions Zn sur l’agrégation du peptide et sur la production d’ERO. D’abord, l’agrégation du peptide est modulée par les ions Cu et Zn. Elle est également très dépendante du ratio métal : peptide, du pH, de la température, etc. Dans le cas de l’agrégation du complexe Cu-Aβ, le ratio métal : peptide est très important. En effet, avec des ratios sous-stœchiométriques en ions Cu comparés au peptide, des fibres sont formées, alors qu’avec des ratios au moins stœchiométriques, ce sont les oligomères toxiques qui sont formés. Les ions Zn impactent également l’agrégation du peptide, même à des stœchiométries très faibles. A des ratios stœchiométriques, des fibres sont formées mais seraient différentes des fibres apos. L’agrégation en présence des ions Cu(II) et Zn(II) serait la même que celle en présence de Zn(II) : les ions Zn, semblent imposer la morphologie des espèces formées. La production de ROS en fonction de ces deux ions métalliques est également détaillée. Cependant, il n’y a à ce jour pas suffisamment d’études pour proposer une tendance. Certains proposent une diminution de la production d’ERO en présence de Zn(II) alors que d’autres ne voient pas d’effet. Ensuite, la production d’ERO par les espèces agrégées est rapportée. A nouveau, il n’y a pas assez d’études pour pouvoir affirmer une tendance, mais il

~ 26 ~

Résumé en français semblerait que les espèces agrégées produisent moins d’ERO que les espèces monomériques. De plus, la présence de Zn semblerait diminuer les ERO produites par les agrégats.

La dernière partie de cette revue s’est concentrée sur l’impact du Zn(II) sur la chélation des ions Cu. Un ligand dans le contexte de la MA doit avoir une constante d’affinité pour le Cu supérieure à celle pour Aβ, mais pas trop forte non plus, pour ne pas retirer les ions Cu non toxiques d’autres métalloprotéines. Puisque les ions Zn ont un impact sur la coordination du Cu, sur la production d’ERO ainsi que sur l’agrégation du peptide, il est pertinent d’étudier son impact sur la chélation des ions Cu.

Les premières études ont été réalisées avec les métallothionéines Zn7-MT-3 puis MT-2A. Il se produit un échange d’ions métalliques Cu et Zn entre les MT et le peptide Aβ. Il se forme un cluster Cu(I) 4- thiolate stable à l’air, stoppant ainsi la production d’ERO. Ensuite, ces études ont été réalisées avec des ligands synthétiques L2 et Lc. Les deux ligands, ayant une constante d’affinité pour le Cu(II) supérieure à celle de Aβ, retirent Cu(II) du peptide et diminuent voire stoppent la production d’ERO. Cependant, en présence de Zn(II), seul le ligand L2 peut toujours retirer Cu(II). La thermodynamique permet d’expliquer ceci. En effet, la sélectivité, c’est-à-dire la différence entre la constante d’affinité pour le Cu et celle pour le Zn, est un paramètre important dans la chélation des ions Cu. Elle est supérieure à celle d’Aβ dans le cas de L2 alors qu’elle est inférieure dans le cas de Lc. Il est donc nécessaire d’avoir une constante d’affinité pour Cu supérieure à celle d’Aβ mais également une sélectivité (Cu versus Zn) supérieure à celle d’Aβ.

Ainsi, cette revue décrit les interactions mutuelles entre les ions Cu et Zn et le peptide Aβ. Zn(II) déplace le Cu(II) de son site de coordination, a un impact sur la production d’ERO ainsi que sur l’agrégation. Il impacte également la chélation du Cu(II), pouvant aller jusqu’à l’empêcher. Enfin, puisque le Zn(II) a un impact sur la chélation du cuivre, il faudrait également étudier l’impact d’autres ions biologiquement pertinents tels que les ions Ca(II) ou Fe(II/III) sur la chélation des ions Cu dans le cadre de la MA.

III-B. Effet thermodynamique de la présence des ions Zn

La première étape de la chélatothérapie contre la MA a été de comprendre et d’étudier comment un ligand pouvait retirer le Cu(II) du peptide Aβ et stopper la production d’ERO. De nombreuses études ont proposé des ligands organiques et peptidiques dans ce sens. Cependant, jusqu’à ce jour, les quelques ligands qui ont pu arriver en essais cliniques ont échoués. Ceci serait surement dû au fait que les ions Cu sont dans un environnement in vivo trop simplifié dans les études in vitro. Par exemple, la forte concentration en ions Zn n’est pas représentée dans ces études. Les ligands comme le clioquinol peuvent également chélater les ions Zn(II) avec une forte affinité, entrainant ainsi une possible

~ 27 ~

Résumé en français chélation in vivo des ions Zn(II) et laissant les ions Cu toxiques dans la fente synaptique. De plus, il a été montré que les ions Zn interfèrent dans la coordination des ions Cu au peptide, dans la production d’ERO par le complexe Cu-Aβ et sur l’agrégation de ce complexe peptidique. Ainsi, la seconde étape de ces études est de mimer cet environnement riche en Zn(II) et de déterminer son impact sur la chélation des ions Cu. Un système à quatre partenaires, c’est-à-dire Aβ, le ligand, le Cu(II) et le Zn(II), est ainsi étudié.

Rappelons qu’un « bon » ligand a besoin d’une constante d’affinité pour les ions Cu supérieure à celle d’Aβ pour pouvoir retirer les ions Cu du peptide. Cependant, en présence de Zn(II), ceci n’est plus suffisant. Deux ligands, L2 et Lc (Figure III-2), sont étudiés pour illustrer ce problème thermodynamique. Les deux ont une affinité pour les ions Cu et Zn supérieure à celle d’Aβ et les deux retirent les ions Cu(II) du peptide en absence de Zn(II).

Figure III-2. Schéma des ligands L2 et Lc.

Avant de regarder l’impact de ces ligands sur la production d’ERO et sur l’agrégation, il est nécessaire d’étudier l’impact que peut avoir le Zn(II) sur ce retrait du Cu(II) d’Aβ. Pour cela, des études spectroscopiques par UV-visible, RPE et XANES ont été réalisées. Concernant le L2, la présence de Zn(II) n’a pas d’impact sur la chélation des ions Cu. Cependant, Lc n’est plus capable de chélater Cu(II) en présence de Zn(II) ; le Cu(II) reste lié au peptide, et Lc coordine Zn(II). Ceci semble contre-intuitif puisque les deux ligands ont une constante d’affinité pour Cu(II) supérieure à celle d’Aβ. Une explication thermodynamique est proposée. En effet, la différence qui existe entre ces deux ligands est leur sélectivité. La sélectivité est le rapport entre la constante d’affinité pour le Cu(II) et celle pour le Zn(II). Notons qu’une forte constante d’affinité pour le Cu(II) n’implique pas une grande sélectivité et inversement, une grande sélectivité n’est pas forcément due à une forte constante d’affinité pour le Cu(II). L’important est vraiment le rapport entre l’affinité du Cu(II) et celle du Zn(II). Ainsi, pour retirer le Cu(II) d’Aβ en présence de Zn(II), le ligand a besoin non seulement d’une affinité pour Cu(II) supérieure à celle d’Aβ mais aussi une sélectivité supérieure à celle d’Aβ. Cette sélectivité pour Aβ est déjà élevée, 104.2 à pH 7.1. L2 a une sélectivité supérieure à celle d’Aβ alors que pour Lc, elle est

~ 28 ~

Résumé en français inférieure, 107.7 et 102.0 respectivement. Ainsi, thermodynamiquement, seul L2 est capable de retirer le Cu(II) d’Aβ en présence de Zn(II).

La deuxième partie de l’étude s’est centrée sur l’efficacité de ces deux ligands à arrêter la production d’ERO. Ceci a été réalisé via deux expériences différentes : le suivi de la consommation d’ascorbate par production d’ERO par spectroscopie UV-Vis (Figure III-3) et le suivi de la cinétique de fluorescence du 7-OH-CCA, molécule fluorescente formée suite à la réaction entre le CCA et le radical HO•. En absence de Zn(II), les deux ligands sont capables de réduire très fortement voire stopper la production d’ERO. Cependant, en présence de Zn(II), seul L2 arrête toujours la production d’ERO. Ceci s’explique par le fait qu’en présence de Zn(II), Cu(II) est chélaté par L2 capable d’arrêter la production d’ERO, alors que Cu(II) est chélaté par Aβ dans le cas de Lc, catalysant ainsi la production d’ERO.

Figure III-3. Suivi de la consommation d’Ascorbate à 265 nm avec une correction de la ligne de base à 800 nm. Le t0 des expériences a été arbitrairement décalé pour une meilleure visibilité. (a) Cu; (b, b') Cu(II)-Aβ ou Cu(II),Zn(II)-Aβ, (c) Cu(II)-Aβ + L, (d) Cu(II)-Aβ + Zn(II)-L, avec L = L2 (panneau A) et L = Lc (panneau B). L’Ascorbate est ajout en premier, et le ligand est ajouté à A = 0.8. [Aβ] = [L] = 12 µM, [Zn] = 12 µM; [Cu] = 10 µM, [Asc] = 100 µM, [HEPES] = 0.1 M, pH 7.1, T = 25°C.

La dernière étude de cette partie concerne l’agrégation du peptide Aβ. Elle a seulement été réalisée pour L2 et non pas pour Lc puisque en présence de Zn(II), le Cu(II) est chélaté par Aβ et donc la présence de Lc n’a pas d’impact sur l’agrégation. Le suivi de l’agrégation a été réalisé par suivi de fluorescence de la ThT et les échantillons ont été imagés par AFM. Rappelons que les conditions utilisées dans cette thèse permettent d’observer que le peptide apo ainsi que le complexe Zn(II)-Aβ agrègent en fibres, alors que Cu(II)-Aβ agrège en oligomères ou protofibres. Les résultats obtenus montrent que si L2 est ajouté à Cu(II)-Aβ, alors l’agrégation est la même que celle de Aβ apo. De plus, si Zn(II)-L2 est ajouté à Cu(II)-Aβ, l’agrégation est la même que celle de Zn(II)-Aβ. Ainsi, même en présence de Zn(II), l’ajout de L2 empêche la formation des oligomères qui sont définis comme étant les espèces les plus toxiques de l’agrégation du peptide. Cette étude prouve à nouveau qu’il y a un échange des ions métalliques entre Aβ et L2.

~ 29 ~

Résumé en français

Cette troisième preuve de concept propose un nouveau critère pour la conception de ligand pour la chélatothérapie des ions Cu dans le cadre de la MA : la sélectivité des ions Cu versus Zn(II). En effet, il est important de prendre en considération la présence des ions Zn(II) en haute concentration dans la fente synaptique. L’étude des ligands L2 et Lc a permis d’illustrer ce problème. Ainsi, un ligand a besoin d’avoir une sélectivité, c’est-à-dire un rapport entre l’affinité pour Cu et celle pour Zn, supérieure à celle d’Aβ pour pouvoir retirer le Cu(II) du peptide dans un environnement riche en Zn(II). Ainsi, L2 arrête la production d’ERO par Cu-Aβ et empêche le Cu(II) de stabiliser des oligomères toxiques.

Cette étude est un premier pas vers une complexification du système. En effet, les ions Zn(II) ont été ici pris en compte pour la chélation des ions Cu. Cependant, il serait intéressant d’étudier également l’impact des autres biomolécules présentes dans les fentes synaptiques ainsi que les autres ions métalliques. De plus, ce concept peut également être appliqué pour la chélation des ions Cu(I).

III-C. Le concept du «pull-push»

Le concept du «pull-push» se place dans le contexte de la chélatothérapie des ions Cu contre la MA. Si les ligands ont une constante d’affinité pour les ions Cu inférieure à celle d’Aβ, ils ne pourront pas retirer le Cu du peptide ; une affinité du même ordre de grandeur que celle d’Aβ entraine un équilibre entre les espèces Cu-Aβ et Cu-L ; une affinité supérieure à celle d’Aβ permet un retrait total du Cu d’Aβ. Rappelons que ceci est vrai en absence de Zn. La présence de Zn, comme expliqué précédemment, perturbe ces équilibres. Il est nécessaire que la sélectivité du ligand soit supérieure à celle d’Aβ pour qu’il puisse retirer Cu d’Aβ (Table III-1, L >> Aβ). Ainsi, un ligand ayant une affinité de l’ordre de celle d’Aβ, ne retire qu’environ la moitié du Cu d’Aβ, alors qu’en présence de Zn, la totalité du Cu peut être retirée par le ligand si la sélectivité est suffisamment grande. Aβ attire le Zn(II) qui à son tour pousse le Cu(II) dans le ligand. Voici quelques exemples pour illustrer ce concept du «pull- push». Considérons les équilibres suivants.

Equilibre 1:

−+↔+ − , avec

. − = = − .

~ 30 ~

Résumé en français

Equilibre 2:

− + − ↔ − + − , avec

− . − . = = − . − .

représente la constant d’affinité de L’ pour M, avec L’ = Aβ ou L et M = Cu ou Zn.

Les équilibres 1 et 2 décrivent donc l’équilibre de retrait du Cu d’Aβ par L en absence et en présence de Zn, respectivement. La Table III-1 montre différents exemples d’affinité et de sélectivité ainsi que leur impact sur les équilibres 1 et 2, à pH 7.1.

Table III-1. Table montrant différents exemples différents exemples d’affinité et de sélectivité d’un ligand et leur impact sur les équilibres 1 et 2.

Constante K % Cu(II)-L Sélectivité de L K % Cu(II)-L d’affinité de L 1 2 101 10-3 3 % L >> Aβ 102 90 % 104 1 50 % 1011 >> 109 107 103 97 % 101 10-3 3 % L = Aβ 1 50 % 104 1 50 % 109 = 109 107 103 97 % 101 10-3 3 % L << Aβ 10-2 10 % 104 1 50 % 107 << 109 107 103 97 %

Le concept du «pull-push» concerne deux catégories de ligands. La première correspond aux ligands avec une affinité pour le Cu de l’ordre de celle d’Aβ (109 M-1 à pH 7.1) (Table III-1, L = Aβ) et une sélectivité très supérieure à celle d’Aβ (~ 104). Ainsi, le ligand doit avoir les affinités suivantes :

-1 -1 = 10 M et = 10 M à pH 7.1 pour pouvoir passer de 50 % de Cu(II) chélaté par L en absence de Zn(II) à 97 % en présence de Zn(II). La seconde correspond aux ligands qui ont une affinité pour le Cu(II) très faible (Table III-1, L << Aβ). Notons que si l’affinité pour Cu est faible, il faut une affinité pour le Zn très petite, sachant que 101-2 est déjà très faible, pour avoir une sélectivité relativement forte (le cas de l’affinité à 107 et de la sélectivité à 107 est très probablement impossible

~ 31 ~

Résumé en français

à avoir). Cette catégorie de ligands permet de passer de 10 % de Cu chélaté au ligand en absence de Zn(II) à 50 % en présence de Zn(II).

Pour illustrer le concept du «pull-push», trois ligands ont été étudiés (Figure III-4). Les constantes d’affinité pour le Cu(II) de ces ligands sont 3.2 109 M-1, 3.0 108 M-1 et 3.7 109 M-1 à pH 7.1 pour L, ABH et BAH respectivement. Les constantes d’équilibre en absence de Zn valent respectivement 2.0, 0.2 et 2.3. Ceci signifie qu’en absence de Zn, L et BAH retirent ~ 60 % d’Aβ et ABH 30 % de Cu(II) d’Aβ. Les constantes d’affinité pour le Zn(II) ne sont pas connues, ainsi il n’est pas possible de déterminer les constantes de l’équilibre 2. Différentes expériences spectroscopiques ont été réalisées pour illustrer la première catégorie de ligands « pull-push ».

Figure III-4. Schéma des trois ligands L, ABH et BAH utilisés pour le «pull-push». B correspond à une β-alanine.

Des études par RPE ont été réalisées (Figure III-5). Une compétition entre Aβ, les ligands, le Cu(II), avec ou sans Zn(II) a été étudiée. Les résultats montrent qu’en absence de Zn, L retire 60 % de Cu(II) d’Aβ, ABH 25 % et BAH 50 %. Ceci est en accord avec les données thermodynamiques. En présence de Zn, L retire 80 % de Cu(II) d’Aβ, ABH 45 % et BAH 70 %. Ces expériences montrent bien l’effet «pull- push» : le Zn tire le Cu hors du peptide et le pousse dans le ligand.

Figure III-5. Expériences RPE de compétition entre Aβ et L (Panneau A), ABH (Panneau B) ou BAH (Panneau C). (a) Aβ + Cu(II), (b) L* + Cu(II), (c) Aβ + Cu(II) + L*, (d) Aβ + Cu(II) + Zn(II)-L*. (L* = L, ABH ou BAH). [L*] = [Aβ ] = [Zn(II)] = 200 μM, [65Cu(II)] = 190 μM, [HEPES] = 50 mM. pH = 7.1. T = 110 K. 10 % de glycérol est utilisé comme cryoprotecteur. Le temps d’incubation des échantillons pour L et ABH est de 64 h au frigo et 7 h à température ambiante pour BAH.

~ 32 ~

Résumé en français

Ensuite une étude par UV-Vis a également permis de mettre en évidence l’effet «pull-push» avec le ligand L. En effet, seul L a pu être étudié par UV-vis car il est le seul à avoir une intense bande d’absorption en présence de Cu(II) à 330 nm. Ainsi, le complexe Cu(II)-Aβ est formé puis L est ajouté. Lorsque l’équilibre thermodynamique est atteint, un équivalent de Zn(II) est ajouté, et ainsi de suite jusqu’à 5 équivalents de Zn(II). Les valeurs d’absorbance de Cu(II)-L en fonction du nombre d’équivalents de Zn(II) sont regroupées dans la Table III-2, ainsi que les pourcentages de Cu(II) lié à L. Notons qu’en absence de Zn(II), la quantité de Cu(II)-L est un peu élevée comparée aux données thermodynamiques, peut-être à cause d’une quantité initiale en L plus importante que prévu. La quantité de Cu(II)-L augmente à chaque ajout de Zn(II), pour atteindre 100 % du Cu(II) lié à L en présence de 5 équivalents de Zn(II). Cette expérience par UV-Vis permet de bien pouvoir visualiser le concept du «pull-push».

Table III-2. Tableau regroupant les valeurs d’absorbance à 330 nm de Cu(II)-L en fonction du nombre d’équivalents de Zn(II) ajoutés ainsi que les pourcentages de Cu(II)-L correspondants.

Nombre 0 1 2 3 4 5 d’équivalents Abs (330 nm) 0,28 0,30 0,32 0,34 0,35 0,36 % Cu(II)-L 78 % 83 % 89 % 94 % 97 % 100 %

Cette dernière preuve de concept repose donc sur l’utilisation des ligands de relativement faible affinité pour le Cu(II), de l’ordre de celle d’Aβ, pour retirer le Cu(II) d’Aβ. Cependant, il n’est possible de retirer qu’environ la moitié du Cu(II). La présence de Zn(II) est ici primordiale : le Zn(II) va tirer le Cu(II) hors du peptide et le pousser dans le ligand. Notons que pour cela, le ligand doit avoir une sélectivité bien supérieure à celle d’Aβ. Trois ligands ont été étudiés pour illustrer ce concept. Un équivalent de Zn(II) permet d’augmenter la quantité de Cu(II)-ligand. Concernant L, 5 équivalents de Zn(II) sont nécessaires pour retirer la totalité de Cu(II) d’Aβ. Cependant, in vivo, ceci ne devrait pas poser de problème puisque la concentration en Zn(II) dans la fente synaptique serait 10 à 100 fois supérieure à celle en Cu.

Un ligand «pull-push» dans le cadre de la MA est intéressant. En effet, puisqu’il a une relativement faible affinité pour Cu(II), il ne pourra pas retirer les ions Cu essentiels des métalloprotéines, sauf dans un environnement riche en Zn comme dans les fentes synaptiques. Ainsi, le concept du « pull-push » est une stratégie de prodrogue.

~ 33 ~

Résumé en français

IV. Conclusion

Cette thèse s’est focalisée sur l’étude de nouvelles preuves de concept concernant la chélatothérapie du Cu dans le cadre de la MA. Quatre nouveaux critères à prendre en compte dans la conception de ligands contre la MA sont proposés et regroupés dans la Figure IV-1.

Figure IV-1. Schéma synthétisant les différents ligands et concepts étudiés au cours de cette thèse.

Le premier concerne la cinétique de captation des ions Cu par le ligand (Figure IV-1, rose). Deux séries de ligands macrocycliques sont étudiées. Il en découle que la géométrie du complexe a un impact sur la cinétique de chélation. En effet, un des macrocyles, le cyclam, a une cinétique de chélation du Cu(II) très lente dans nos conditions comparée à celle du cyclen. De plus, le cyclam chélate le Cu(II) dans sa cavité macrocyclique, alors que le cyclen le chélate hors de sa cavité. Ainsi, la géométrie du complexe est un paramètre important pour le design du ligand. De plus, en vue d’améliorer la cinétique de captation, des bras picolinates ont été greffés sur les macrocyles. Les expériences menées ont montré que ces bras accéléraient la captation des ions Cu par le ligand. Ainsi, cette première preuve de concept met en avant l’impact de la cinétique de captation, mais également l’importance du design du ligand.

Le second critère proposé ici est le fait que le ligand puisse cibler à la fois le Cu(I) et le Cu(II) (Figure IV-1, vert). Ceci est important puisque le degré d’oxydation du Cu dans la fente synaptique n’est pas connu. Cependant, un tel ligand a la possibilité de chélater le Cu(I) et le Cu(II), il est donc possible qu’il cycle facilement entre ces deux états redox et donc qu’il produise des ERO. Le ligand étudié dans cette

~ 34 ~

Résumé en français partie est en effet capable de retirer Cu(I) et Cu(II) d’Aβ. Cependant, il ne produit pas d’ERO. Ceci serait dû au fait que les deux coordinations et géométries sont différentes, et donc le passage redox d’un complexe à l’autre est lent. A nouveau, ce critère est important à prendre en compte dans le design de ligands.

Le troisième critère concerne la thermodynamique, notamment l’impact que peut avoir le Zn(II) sur la chélation du Cu (Figure IV-1, orange et violet). En effet, pour qu’un ligand puisse retirer le Cu(II) d’Aβ en présence de Zn(II), il doit avoir une sélectivité, c’est-à-dire un rapport d’affinité entre le Cu et le Zn, supérieure à celle d’Aβ. Dans le cas contraire, comme l’a montré l’étude des deux ligands L2 et Lc, le Cu(II) reste lié au peptide entrainant la toxicité relative à Cu-Aβ.

Enfin, la dernière preuve de concept proposée ici est le concept du «pull-push» (Figure IV-1, rouge). Il s’agit d’un ligand avec une affinité pour le Cu de l’ordre de celle d’Aβ, mais avec une sélectivité supérieure. Trois ligands ont été étudiés pour illustrer ce concept. En absence de Zn(II), le ligand ne peut retirer qu’environ la moitié de Cu(II) d’Aβ. En présence de Zn(II), le ligand retire plus que la moitié de Cu, voire la totalité. Ce concept est intéressant dans la chélatothérapie du Cu contre la MA puisqu’il permet d’utiliser des ligands qui ne peuvent pas retirer le Cu des métalloprotéines, excepté en présence de Zn(II), comme dans les fentes synaptiques.

Ensuite, des études concernant l’impact des ions Zn(II) sur la cinétique de captation ont été commencées. Dans certains cas, les ions Zn(II) sont chélatés par le ligand en premier, et leur vitesse de dissociation étant très lente, empêchent le retrait du Cu(II) d’Aβ. Il serait intéressant d’étudier aussi des ligands permettant d’illustrer la deuxième catégorie de ligands du concept du « pull-push », les ligands avec une faible affinité pour Cu(II) ; de regarder l’impact du « pull-push » avec le Cu(I).

Plus tard, il sera intéressant de combiner ces critères avec ceux déjà existants (en particulier la capacité à passer la BHE, la toxicité intrinsèque du ligand et du complexe) dans un seul ligand et de tester la capacité de ce nouveau ligand dans le cadre de la MA. Sûrement, d’autres critères seront à ajouter pour enfin trouver LE ligand contre la MA. On pourrait penser à l’excrétion du complexe de Cu(II) et donc au passage du complexe de la BHE, mais également à d’autres critères tels que l’impact des autres biomolécules environnantes ou des autres ions métalliques présents dans la fente synaptique.

~ 35 ~

General introduction

General introduction

Alzheimer’s disease (AD) is the most common neurodegenerative disease. Elderly people are the most affected by this disease. Nowadays, there is no known cure and this is an important issue, especially since life expectancy continuously increases. There is an important need to discover drugs against AD. This disease is difficult for the patients but also for their families, friends and caregivers. The patient can exhibit memory losses, troubles in the spatio-temporal framework, troubles in speaking, etc. Sometimes, patients can become aggressive. In the late stage of AD, the deleterious effects on the brain lead to the death of the disease’s sufferers.

In the AD brains, different events happen. This thesis focuses on the formation of the senile plaques, between the neurons, precluding the synaptic connexions. These plaques are mainly composed by the Amyloid-β (Aβ) peptide and metal ions such as Cu and Zn ions. The Aβ peptide aggregates into the senile plaques following different steps. Furthermore, there is a dyshomeostasis of the Cu and Zn ions. The most supported hypothesis is that Cu ion level is too low intra-cellulary, while in excess extra-cellulary. For Zn ions, the tendency is not clear. Another important parameter of AD and under focus in this thesis is the Reactive Oxygen Species (ROS) production catalysed by the Cu- Aβ complex. A too high concentration of ROS is deleterious for the surrounding biomolecules, such as the neuronal membranes. Cu(II)-Aβ can be reduced by a reductant such as ascorbate and re-oxidized by dioxygen, leading to the ROS production.

Different therapeutic approaches against AD exist. Cu chelatotherapy is one of them. The idea is to develop Cu(II) ligands, able to remove Cu ions from Aβ and to stop associated deleterious events (ROS, aggregation, etc.). Many ligands have already been studied, and the most developed ones are clioquinol and PBT2. They went through clinical trials but failed in Phase II. This failure can be due to a lack a selectivity (i.e. the ratio between the affinity constant for Cu(II) and the one for Zn(II)) of Cu over Zn ions: they are able to chelate Cu and Zn ions, with a high affinity constant. This thesis focuses on the Cu chelatotherapy and proposes new concepts regarding criteria that have to be fulfilled by the ligand in order to remove Cu ions from Aβ and stop ROS production.

In the first part of this thesis, general features of AD are described. The clinical signs, the risk factors, the histopathological hallmarks and the different diagnostic tools are presented. Then, the

~ 36 ~

General introduction different interactions between Aβ and the metal ions are detailed. The available symptomatic drugs as well as the therapeutic approaches against AD are reviewed.

In the second part of this thesis, the Cu chelation is under focus. After a description of the different ligands used in chelatotherapy, an unexplored study is shown aiming at answering the following questions: - Does the kinetic of Cu chelation by a ligand have an impact on the ROS production catalysed by Cu-Aβ complex? Note that during the ROS production, Cu oscillates rapidly between Cu(II) and Cu(I) and most of the chelators described up to now target only Cu(II). - How is it possible to avoid that the ligands fail in the removal of Cu ions due to kinetic reasons (despite thermodynamically favoured) when designing a ligand against AD? - Does the complex geometry have an impact on the kinetic of the Cu chelation? Finally, another point is addressed. The redox state of Cu ions in the synaptic cleft is not well known and during the ROS production, Cu ions do redox cycle. - Which redox state of Cu ions should be targeted by the ligand to efficiently remove Cu ions from Aβ? Cu(I), Cu(II) or both? - Is it possible to target both redox states? - If yes, is there a risk that the Cu-complex formed with the ligand itself can produce efficiently ROS by cycling between Cu(I) and Cu(II)? What parameters have to be considered? This part answers to these questions with a ligand able to chelate both redox states.

The last part of this thesis focuses on the impact of Zn ions on the Cu chelation by a ligand from Cu-Aβ complex. Indeed, the concentration of Zn ions in the synaptic cleft is often 10 to 100 times higher than the one of Cu ions in the synaptic cleft. There is a possibility for the ligand to chelate Zn ions instead of Cu ions. Therefore, the “healthy” Zn(II) will be removed from the synaptic cleft where it is essential, while the “toxic” Cu ion will stay bound to Aβ in the synaptic cleft, producing ROS. First, a state of the art regarding the mutual interactions of Cu and Zn ions with Aβ is reported. Coordination, ROS production, aggregation and chelation are the key points of this review of literature. Then, a first study illustrates the importance of thermodynamic equilibria in the Cu chelation therapy and addresses the following issues. In the absence of Zn(II), a higher affinity constant for Cu of the ligand than Aβ is sufficient to remove Cu ions from Aβ (if the kinetic is favourable) - Is it also the case in the presence of Zn ions? Finally, the last part of this section illustrates a last new concept: the “pull push” effect. - Is a ligand with an affinity constant for Cu in the same range than Aβ able to remove totally Cu from Aβ?

~ 37 ~

General introduction

- Can Zn(II) help the ligand to chelate Cu ions from Aβ in this case? Finally, a general conclusion answers these questions and proposes some perspectives with respect to the Cu chelatotherapy against AD.

~ 38 ~

Chapter I: Context of the project

Chapter I: Context of the project

This chapter focuses on the description of Alzheimer’s disease, with the prevalence, the different symptoms, the risk factors, the histopathological hallmarks and the different diagnostic tools. Then this first chapter details one of the hallmarks of the disease, the senile plaques which are made of the Amyloid-β (Aβ) peptide and the interactions of Aβ with different metal ions present in the brain. Later on, the current treatments and therapeutic approaches are reported. Finally, the objectives of the present study are described.

I-A Alzheimer’s disease

I-A.i Prevalence and symptoms

Alzheimer’s Disease (AD) has first been described by Alois Alzheimer in 1906 with the case of a 51- years old woman, dead after four and a half years of illness.1 It is the most common neurodegenerative disease worldwide and represents 50 to 80 % of all dementia. More than 35 million people are affected by the disease and this prevalence is estimated to triple by 2050, due to the increase of life expectency.2 Different stages of the disease have been identified.3 In the early-stage, people are able to work, to drive, to have social relationships, etc. They are still independent. Nevertheless, they have some trouble memories, such as familiar words or the location of everyday objects. The middle-stage, is generally the longest stage of the disease. Patients are suffering from important memory loss. They can have, for example, troubles in the spatio-temporal framework. Difficulties in expressing or performing routine tasks are also common symptoms. In addition, AD persons can have some behavioural changes. In the late-stage, people have difficulties in speaking, walking, or doing everyday actions. Furthermore, their memory and cognitive skills are worsened.

I-A.ii Risk factors

The age of the population is the most important risk factor. The prevalencea of AD is estimated to be 4.4 % by people of 65 years-old and more in Europe.4 Note that this figure has been given in 2000 and from our best knowledge, more recent studies don’t exist. Only figures from the USA appear

a The prevalence is a percentage of the number of cases in the population. b The incidence is the number of new cases in a given period in the population. ~ 39 ~

Chapter I: Context of the project regularly.5 For example, in 2017, in the USA, around 5.5 million people are affected by AD, where 4 % of AD patients are less than 65 years-old, 16 % between 65 and 74 years-old, 44 % between 75 and 84 years-old and 38 % are more than 85 years-old. Note that the total percentage is higher than 100 due to the rounding.5 The annual incidenceb rate of AD is given age by age in the Figure I-1. These figures are alarming due to the increasing life expectancy and the lack of cure.

Figure I-1. Graphic illustrating the annual incidence rate by age of AD for 100 persons-years. Figure from ref.6

It is not well defined whether the gender is a risk factor of AD.7 Indeed, estrogen seems to play a neuroprotective role, with an unknown mechanism.7 After the menopause, woman serum contains less estrogen than man serum, the latter should have more neuroprotectors than woman serum. Some clinical trials of hormone replacement therapy have been performed, with no clear trend. Nevertheless, women are known to live longer than men. Therefore, they rich the “Alzheimer age” risk more often than men, and so they are more exposed to the disease than men (in this case, the age is the risk factor).

Other studies have demonstrated that the prevalence of the disease can decrease with a better way of life.8-10 Indeed, diabetes mellitus, mid-life hypertension, depression, physical inactivity, smoking and cognitive inactivity are also risk factors of AD that can be prevented.8-10 Moreover, the obesity is also a risk factor of Alzheimer’s disease.11 Chronic stress has been described as a potential accelerator of appearance of AD or as an effect that increases the incidence of the disease.12 Hence, chronic stress is considered as an eventual risk factor for the disease.12

Some genetic modifications are also a risk factor of AD. For example, more than half of the AD patients have the APOE ε4 allele rather than the ε2/ε3 on the apolipoprotein E, the protein which transports lipids and which is responsible for the neuronal membrane care and remodelling.13-14 Furthermore, other mutations on the Amyloid Precursor Protein (APP) for example and on Presenilin 1 and 2, which are a subunit of the γ-secretase, can be a risk factor for the disease6, 15 (for more details on APP and γ-secretase, see § I-B.i). Table 1 summarises the different mutations found in AD patients. Other mutations on the Aβ peptide (for more details, see § I-B.i) are familial AD mutations,16 other ones impact the amino acids involves in the metal ion coordination to Aβ.

~ 40 ~

Chapter I: Context of the project

Table 1- Table summarising the main alterations found in AD patients, from ref.6

Gene Main alteration Presumed mechanism

Amyloid precursor protein (APP) Mutation Autosomal dominant, mostly early onset

Presenilin 1 (PSEN1) Mutation Autosomal dominant, mostly early onset

Presenilin 2 (PSEN2) Mutation Autosomal dominant, mostly early onset

Apolipoprotein-E (APOE) Common variant Familial and sporadic, late onset

Sortilin-related receptor, L(DLR class) Common variant Familial and sporadic, late onset A repeats-containing (SORL1)

Clusterin (CLU) Common variant Sporadic, late onset

Phosphatidylinositol binding clathrin Common variant Sporadic, late onset assembly protein (PICALM) Complement component (3b/4b) Common variant Sporadic, late onset receptor 1 (CR1)

Binding integrator 1 (BIN1) Common variant Sporadic, late onset

Some diseases can become a risk factor for AD. Periodontitis, among others, could be linked with sporadic late onset of AD.17 Moreover, Traumatic Brain Injury (TBI) is also an important risk for the development of AD. Indeed, Amyloid β (Aβ) plaques have been found in the brain of 30 % of patient died of TBI.18 Patients with Down’s syndrome, or trisomy 21, have also a high risk to develop AD. This is attributed to the triplication and the over-expression of the APP gene located on chromosome 21.19

I-A.iii Histopathological hallmarks

There are three main histopathological hallmarks of AD. One of them is the loss of mass of the brain.20-21 The size of the brain is diminished particularly in the brain regions involved in memory and in learning (Figure I-2). For example, the hippocampus suffers from a volume reduction of ~ 12 % each year.22 This is due to the degeneration of the synapses and the death of the neurons.

Figure I-2. Comparison between a normal brain (felt) and an AD brain. The loss of mass is easily notable on these pictures. Pictures adapted from ref.21

~ 41 ~

Chapter I: Context of the project

The neurofibrillary tangles of Tau protein are another histopathological hallmark of the disease.23 This protein plays an important role in the stabilization of the microtubules as well as in the microtubule assembly. In the AD brain, this protein is hyperphosphorylated. Tau aggregates in the neuron into pair-helical filaments and then accumulates into neurofibrillary tangles.23 The protein is not able to perform its roles and this leads to the apoptosis of the neurons. A scheme of this hallmark of AD is presented in Figure I-3.

Figure I-3. Scheme representing the different hallmarks of the Alzheimer’s disease. In an AD brain, neurofibrillary tangles of Tau protein and Amyloid plaques are present. Scheme from ref.24

The other important histopathological hallmark of AD is the senile plaques, also called amyloid plaques.25 They mainly consist of the aggregated Amyloid-β (Aβ) peptide (for more details on the Aβ peptide and its aggregation, see § I-B). These aggregates are formed in the extracellular medium, between the neurons (Figure I-3). These plaques preclude the synaptic connection and thus can lead to the neuronal death. The research project presented here focuses on this aspect of the disease: the Aβ peptide and the senile plaques. There is no investigation on the Tau protein in this thesis.

I-A.iv Diagnostic tools

Diagnosing AD is not trivial: this disease has to be differentiated from other dementias. For an early detection, clinical tests are performed on the patients.26-27 It exists the subjective memory complaint (SMC) test, the assessment of the late-onset depression, the speech testing, the olfactory testing, the eye testing, and the gait testing (there is a decrease about 50 % of the gait speed between a healthy people and an AD severe patient).26 A blood test is also important in the diagnostic. Indeed, as about 500 mL of cerebrospinal fluid (CSF) is absorbed in the blood every day, biomarkers of AD could be detected in the blood.26 Many biomarkers such as total cholesterol, plasma Aβ42/Aβ40 ratio, increased expression of inflammatory cytokines, etc., can be detected. The accuracy of blood test diagnostic is quite high: between 70 and 90 % depending on the biomarkers used. If the psychometric tests converge towards AD diagnostic, neuroimaging and CSF analyses are performed.27 Indeed, not all

~ 42 ~

Chapter I: Context of the project of these tests are sensitive for AD but for cognitive disorders in general.26 They are not enough to diagnose AD with precision. For example, if these tests are performed on a neurodegenerative disease patient and on an AD patient, the diagnostic will not be clear on the nature of the dementia. More investigations towards a specific diagnostic are in progress.26

AD Neuroimaging Initiative (ADNI) has begun in 2005.27 The neuroimaging of AD is divided in two main techniques: the Positron Emission Tomography (PET) and the Magnetic Resonance Imaging (MRI).27 The functional MRI techniques are based on the functional integrity of brain networks involved in cognitive domains.28 The structural MRI techniques are based on the progressive brain atrophy and/or on the changes of the tissue involved in AD.28 These techniques allow for example to detect a change in the hippocampus volume or in the ventricular volume, or also in the integrity of the white matter.28-29 Nevertheless, cerebral atrophy is not specific to AD. Therefore, MRI needs a complementary technique, such as PET imaging. PET is a nuclear medicine imaging technique, based on the detection of the gamma ray. A radionuclide emits positrons β+ during its radioactive decay, the positron towards some distance before encountering an electron, causes an annihilation event, emitting two gamma rays in the opposite direction. In the examples given below, the radionuclides are 11C, 18F or 64Cu. There are two main types of PET imaging methods for AD detection: the molecular imaging and the metabolic imaging.27, 29 The first one is used in AD for the detection of Aβ deposits. The first human trial for the detection of Aβ plaques by PET is described by Klunk et al.30 in 2004; it is the Pittsburgh compound B (PiB), a neutral 2-aryl-benzothiazole derivative labelled with the 11C (Figure I-4). PiB compound was designed to bind the Aβ deposits.

Figure I-4. Representation of the Pittsburgh compound B (PiB) used for the Aβ PET imaging.

The Figure I-5 top shows PiB-PET images of a healthy and an AD brains. In the control brain, there is no specific retention of PiB, whereas in the AD brain, there is a high PiB retention. This shows that Aβ deposits are present in the AD brain and not in the healthy brain, as expected. Note that in the regions of the brain which are not involved in AD, there is no retention of PiB both in the control brain and in the AD brain. PiB images should give quantitative information on the amyloid deposits. The limitation of this molecule is the short half-time of its radionuclide which is around 20 minutes. New molecules targeting Aβ deposits,28, 31 such as ¹⁸F-BAY94-9172,32 18F-AV-45,33-34 or also PiB 18F derivatives35-36 are emerging and have the radionuclide 18F which has a longer life-time around 1.8 h. More recently, molecules targeting amyloid deposits with the radionuclide 64Cu were developed.37 The

~ 43 ~

Chapter I: Context of the project half-time of the 64Cu is much higher than the other two ones, around 12.7 h facilitating the PET imaging. Note that molecules targeting Tau deposits are currently under investigations as well.28-29, 31 Other 64Cu complexes for PET imaging are also developed for the detection of Cu ion deficiency, because Cu dyshomeostasis has been linked to AD.38-39 The metabolic imaging uses the Fluorodeoxyglucose PET (FDG-PET).21, 28-29 It is a glucose molecule with a ¹⁸F. The difference between both brains shown in Figure I-5 (bottom) regarding the 18FDG uptake is tremendous: there is an important hypometabolism in the case of the AD brain. This technique is very sensitive,29 but not specific enough. Indeed, the glucose retention in the brain can be altered by many phenomena. Therefore, performing PiB PET and 18FDG PET allows a more accurate diagnostic of AD.

Figure I-5. PiB-PET images (top) and FDG-PET images (bottom) of a 67-years old healthy person as a control and of a 79-years old AD patient. A three-day delay exists between PiB and FDG-PET images. For the PiB-PET standardized uptake value (SUV) images (top), the images are summed over 40 to 60 minutes. In the healthy brain, there is no retention of PiB, contrary to the AD brain. This means that the healthy brain does not expose Aβ aggregates whereas the AD brain does. For the FDG-PET images (bottom), regional cerebral metabolic rate for glucose (rCMRglc) images (in mol.min-1.100 ml) show a normal 18FDG uptake for the healthy brain and a hypometabolism in the case of the AD brain (arrows). Figure from ref.30

Another technique can be performed: the CerebroSpinal Fluid (CSF) assay.27, 29, 40-41 It consists in detecting CSF biomarkers such as Aβ42, Tau and phosphorylated Tau. Contrary to the PET imaging in which accumulation of Aβ could be detected, in the CSF assay, a decrease in the Aβ concentration is observed due to its aggregation in the synaptic clefts.42 But CSF assay could be less reliable because Aβ levels can fluctuate.29 The Aβ accumulation biomarkers used in PET imaging can also be used in the CSF assay. Biomarkers for neuronal injury or neurodegeneration are also developed.42

In brief, after the psychometric tests, the CSF assay, the MRI, the PET targeting Aβ such as PiB-PET, the Cu-PET and the FDG-PET are complementary tools and allows a quite accurate diagnostic of AD.31

~ 44 ~

Chapter I: Context of the project

I-B The Amyloid-β peptide and metal ions

I-B.i Aβ peptide

The Aβ peptide derives from the Amyloid Precursor Protein (APP).25, 43-45 APP is a transmembrane protein with 695 amino acids in its predominant form.44 The physiological role of the APP is still unclear. It was first described as a receptor at the cell-surface, but it is now considered as a cell adhesion biomolecule, involved in neuronal development and synaptogenesis.46 APP can be cleaved by three enzymes. In the amyloidogenic pathway, APP is first cleaved by the β-secretase and then by the γ- secretase (Figure I-6). The residual extracellular peptide is the Aβ peptide.25, 43-45 There is also a non amyloidogenic pathway, involving the α-secretase, avoiding the formation of the full length Aβ peptide.44

Figure I-6. Scheme representing the formation of the Aβ peptide. The APP protein is cleaved by two secretases, leading to the extracellular Aβ peptide. The sequence of the Aβ peptide is written in pink.

The Aβ peptide is a 38 to 43 amino acid residue sequence.25 The first sixteen N-terminal amino acids form the hydrophilic part which is responsible for the metal ion coordination. The amino acids in the hydrophobic part of the peptide are responsible for the aggregation further leading to the so-called senile plaques. Note that the Aβ peptide can exhibit some mutations linked to the familial early-onset AD case,25 and some truncations such as Aβ4-x or Aβ11-x.47

~ 45 ~

Chapter I: Context of the project

I-B.ii Metal ions

Many metal ions are present in the living beings. This study focuses only on the Cu and Zn ions. Copper is an essential metal ion.48 In the human body, Cu is the catalytic centre in many enzymes.48-50 For example, it is the catalytic centre of the Superoxide Dismutase, a (Cu, Zn) enzyme responsible for the regulation of the oxidative stress, via dismutation of the superoxide. Cu ions can also be found in proteins for their transport to the catalytic enzyme or also for their storage. Principally, Cu ions are involved in electron transfer processes or in the binding / activation / reduction of O2 in human bodies.49 On the other hand, Cu ions, due to their redox ability, are able to produce Reactive Oxygen Species (ROS).51 ROS are very deleterious for the surrounding biomolecules and when they are over- produced, they contribute to oxidative stress. This is the reason why Cu ion concentration is tightly regulated. Many Cu-proteins exist: specific transporters which stabilize one redox state of Cu or the other as for example hCTR152 and ATP7A53-54 involved in the Cu uptake and release into and from human neurons (Figure I-7); specific metal ion delivery proteins such as metallochaperones which protect Cu ions from scavengers, etc.51 Cu ions play also an important role in neurotransmission.55 Their tight regulation is also required for the neuronal health. In AD brains, the Cu homeostasis is deregulated.56-60 Cu levels in the CSF are higher,61 while they are lower in the hippocampus and in the intracellular medium55, 62 in AD brains compared to healthy brains. This Cu ion dyshomeostasis may be one of the key event of AD.

Zn(II) is also an essential ion. It plays many roles, including regulatory, structural and enzymatic fonctions.63 Moreover, it is the most abundant “trace metal” in the brain.60 Its concentration in the synaptic cleft is high but not precise, depending the synaptic transmissions. It can reach 10 to 100 µM during its release from the vesicles into the synaptic cleft.56 Zn homeostasis is well regulated, by several Zn transporter families. Intracellular concentrations of Zn(II) are decreased by the ZnT family whereas the ZIP family brings Zn ions from the extracellular medium to the cytoplasm (Figure I-7).60, 64 Zn(II) is a modulator of the neurotransmittion. For example, it is released in the synaptic cleft, modulating receptors for the regulation of the activity of the glutamatergic synapses, like Cu ions.59-60 In AD brains, there is also a Zn(II) dyshomeostasis. 56, 59-60 However, the tendency of the modification of Zn(II) concentrations is not clear: both decreased and increased concentrations are found in the hippocampus, in the serum and also in the CSF of AD patients.60 Hence Zn(II) dyshomeostasis is linked to AD, but it is not clear how and to which extent.

~ 46 ~

Chapter I: Context of the project

Figure I-7. Scheme illustrating the Cu and Zn ions regulation/uptake in neurones.

I-B.iii Interaction between Aβ and the metal ions Cu(I/II) and Zn(II)

As previously explained, Aβ peptide and dyshomeostasis of metal ions such as Cu(II), Cu(I) and Zn are linked to the aetiology of AD. The study of their interaction is thus biologically relevant. Indeed, metal ions have been detected in the senile plaques at high concentration: about 1 mM of Zn and about 400 µM of Cu ions and Aβ contains amino-acid residues able to bind metal ions.65

I-B.iii.1 Coordination and affinity constants

The first step in the study of the interaction between Aβ and Cu and Zn ions is the metal ion coordination and the determination of their affinity constants. Note that these studies have been mostly performed on the monomeric and C-terminally truncated Aβ. Indeed, as explained previously, the metal ion coordination involves the first sixteen N-terminal amino acids of Aβ. Therefore, the Aβ16 encompassing the first sixteen amino acids is a good model58 and is used for these investigations.

Cu(II) coordination to Aβ monomers has been studied by many groups and reviewed.58, 60, 66-68 Many spectroscopic techniques have been used. Depending on the pH, different Cu(II) coordination sites exist. Only those present around the physiological pH are described here. Figure I-8 (top) describes the main components of Cu(II)-Aβ near physiological pH. In Component I, Cu(II) has five ligands: the N-terminal amine, the O atom from the first peptidic bond and two N atoms from His6 and His13 or His14. There is a dynamic exchange between His13 and His14. On the apical position, there is one carboxylate group from the side chain of Asp1, Asp7, Glu3 or Glu11, with a preference for Asp1 via a H-bond and a water molecule.69 In Component II, the five ligands are: the N-terminal amine, the

~ 47 ~

Chapter I: Context of the project

N- from the first peptidic bond, the O atom from the second peptidic bond and a N atom of an imidazole side chain from one of three His residues. There is also a dynamic exchange between those three His. On the apical position, one of the 4 carboxylate groups from the side chain of Asp1, Asp7, Glu3 or Glu11 completes the coordination sphere of Cu(II). At pH 7.4, the affinity constant for Cu(II) of Aβ is about 1010 M-1.70-72 The evaluation of this affinity constant has been investigated by different groups. This is an important parameter for the design of ligands for chelation therapy approach because they need an affinity constant for Cu(II) higher than Aβ. Many works had been carried out and the affinity constant values ranged from 106 M-1 to 1019 M-1 at pH 7.4. But more recently, a consensual value was proposed near 1010 M-1 at pH 7.4. Most of these studies used potentiometric titrations, isothermal calorimetry, and fluorescence or competition experiments.

Figure I-8. Representations of the different metal ion coordination to the Aβ peptide at physiological pH 7.4. The affinity constants are given at pH 7.4. Figure adapted from ref.73

During this PhD work, a competition experiment with an UV-Visible competitor L, a 3,4- bis(oxamato)benzoic acid ligand, has been investigated.74 This technique is easy to handle and to perform due to the intense absorption of Cu(II)-L at 330 nm, and it rapidly gives a value for the affinity constant of the target molecule. Once the affinity constant for Cu(II) of the used competitor L has been determined, the affinity constant of Aβ for Cu(II) can be measured by a competition for Cu(II) between Aβ and L. Using a home-made fitting of the competition data, an affinity constant of Aβ for Cu(II) of 1.6 x 109 M-1 at pH 7.1 has been obtained in line with value obtained by potentiometry by another group71

~ 48 ~

Chapter I: Context of the project and in line with the value of 1010 M-1 reported at pH 7.4. Then, the affinity constant for the entire peptide Aβ40 is determined by the same way, and the value is closed to the Aβ16 peptide. The same experiment has been performed on a wide series of Aβ16 mutants; the mutations focusing on the amino acids involved in the Cu(II) coordination. The affinity constants obtained for each mutant are in line with the Cu(II) binding site of Aβ determined by spectroscopic studies. These experiments also show that L is a good competitor to determine the affinity constant of peptides or proteins with a moderate affinity constant for Cu(II). This study is given in Annexe A of this manuscript.

Cu(I) coordination to Aβ peptide has also been studied by several groups.75-79 All converge towards a linear coordination (see Figure I-8, bottom left) between two His residues. The predominant form involves the N atoms from His13 and His14, in equilibrium with the coordination by His6 and His13 or His6 and His14. The affinity constant of Aβ for Cu(I) has also been determined by competition experiment assay with Ferrozine (Fz).72, 80-81 Two values at physiological pH are found: 1010.4 M-1 72, 81 and 106.9 M-1.80 Such difference is due to the two different values of the affinity constant of the reference Cu(I)(Fz)2 used for the competition studies. More studies have to be performed to determine the precise value of the affinity constant for Cu(I) of Aβ.

Regarding the Zn(II) coordination by Aβ peptide, different models have been proposed and in most of them, the N-terminal amine was involved.82-84 Nevertheless, our recent multi-technical study85 demonstrates that the N-terminal amine is not in the Zn(II) coordination sphere at pH 7.4 (see Figure I-8). This study has been divided in two main parts. The first one describes an EXAFS study paralleled to a NMR study used for the determination of the number and the nature of the ligands. Zn(II) is a 4- coordinated ion in the Aβ peptide, with a tetrahedral geometry. Then, XANES and NMR investigations have been performed on a wide series of Aβ16 mutants in order to determine which amino acid residue is involved in the coordination of Zn(II). If the complex Zn(II)-mutant exhibits the same XANES and NMR signals than the Zn(II)-Aβ complex, the amino acid mutated is not involved in the coordination, whereas if the signals are different, the amino acid mutated can be involved in the coordination. The ligands involved in the Zn(II) coordination are the N atom from His6, the carboxylate from the Glu11, a N atom from His13 or His14 and a carboxylate from Asp1, Glu3 or Asp7. This study is given in Annexe B of this manuscript.

The affinity constant of Aβ for Zn(II) has also been determined and the value is about 105 M-1 at pH 7.4.86-87

~ 49 ~

Chapter I: Context of the project

I-B.iii.2 ROS production

As previously explained, Cu ion homeostasis is essential for health. Indeed, a too weak concentration is lethal as well as a too high concentration. In the AD brains, there is a dyshomeostasis of Cu ions. Aβ peptide is able to coordinate both redox state of Cu ions: Cu(II) and Cu(I). Cu ions are able to catalyse ROS production. ROS are the products of the incomplete reduction of dioxygen by a

88-91 •- 92 reductant, a biologically pertinent being ascorbate. First, superoxide (O2 ) is produced, then

88-91 • 88-91 hydrogen peroxide (H2O2) and then hydroxyl radical (HO ). ROS are deleterious species, due to their high reactivity. Protein oxidation and lipid peroxidation are well known damages of this oxidative stress.93-96 Figure I-9 reminds this ROS production and some methods used for their detection in vitro.

Figure I-9. Scheme representing the ROS production catalysed by the Cu-Aβ complex. The different detection methods are also reminded. Scheme from 73.

In this work, only Ascorbate consumption and 3-Coumarin Carboxylic Acid (CCA) assays were performed. The first assay consists in following the consumption of the Ascorbate, which exhibits an intense absorption at 265 nm (ε = 14 500 cm-1.M-1). The decrease in the concentration mirrors the ROS formation. The CCA assay consists in the detection of the HO•.97 Hydroxyl radical reacts with the CCA, forming the 7-OH-CCA which is a fluorescent molecule. Under an excitation at 390 nm, 7-OH-CCA emits a fluorescence light at 450 nm. The detection of such a fluorescence is correlated to the HO• production by the Cu-Aβ complex and O2 in reductive conditions.

~ 50 ~

Chapter I: Context of the project

I-B.iii.3 Aggregation

The senile or amyloid plaques are one of the hallmarks of the disease. They are mainly present in or around the synaptic clefts of glutamatergic synapses of the cortex and of the hippocampus,60 the loci of the highest Cu concentrations in the brain.60 As previously explained in part I-A.iii, these plaques are mainly composed by the Aβ peptide and contain other molecules and metal ions such as Cu and Zn ions. Aβ is a monomeric and soluble peptide in healthy brains, and aggregates in AD brains. This aggregation process is part of the so-called amyloid cascade hypothesis.98-103 Nowadays, this hypothesis is one of the most accepted, although still discussed.104-107 The amyloid cascade hypothesis describes the aggregation of Aβ as the central event of the disease. This aggregation is an auto-catalytic self-assembly phenomenon.108 Figure I-10 illustrates the aggregation process of the Aβ peptide. The kinetic of Aβ aggregation is mathematically described by a sigmoidal curve.47, 108 The following equation

109 is the model for this kinetic: , where F0 is the baseline level before the / aggregation, k is the elongation rate constant, A is the amplitude and t1/2 is the time at half of the aggregation process.

Numberof fibrils

Figure I-10. Scheme describing the aggregation process and the ThT fluorescence kinetic shape for the associated experiment. Scheme adapted from 110.

The nature of the species during the aggregation process changes. The nucleation phase during which low molecular weight species such as oligomers are formed is the first phase of the aggregation. Note that oligomers have proposed to be the most toxic species of the aggregation due to their size and interaction with the membranes.111-112 Then, there is the elongation phase during which the oligomers elongates into protofibrils and fibrils. The “plateau” or equilibrium phase is the

~ 51 ~

Chapter I: Context of the project thermodynamic equilibrium. 47, 108 At a molecular level, the hydrophilic N-terminal part of the peptide may be involved in the aggregation process but remains mainly disordered in the fibrils. The 12 to 24 amino acid residues and 30 to 40 amino acid residues adopt β-strand conformation with electrostatic interaction between D23 and K28 in the loop of the β-strand.113 Interaction between β-strands has been observed forming the β-sheet structure, i.e. there are hydrogen bonds between the peptide bonds of the hydrophobic part of the peptide along the fibril axis (β-sheet). It also exists interactions between the peptidic side chains perpendicular to the fibril axis.108, 113 Note that in this thesis, the term aggregates defines either amorphous species or fibril species (i.e. amyloid species made of β-sheets).

Metal ions such as Cu and Zn ions have an impact on the aggregation of the peptide.58, 108, 114 With the working conditions used during this thesis, Cu ions stabilize oligomers and other small aggregates which are proposed to be the most toxic species, while Zn ions lead to the fibril formation. The aggregation process depends on many conditions that is why different groups find different results.

In this work, in order to follow the aggregation of Aβ, a classical fluorescence assay is used.115 The Thioflavin T, called ThT (see Figure I-11), is a fluorophore with a free rotation of the bond between the two aromatic moieties. ThT is able to interact with β-sheets, leading to preclusion of the free rotation. This phenomenon enhances its fluorescence. Therefore, during an aggregation experiment, the formation of β-sheets is monitored by the detection of the fluorescence. Hence, ThT fluorescence is used to monitor fibril formation. A particular caution is needed to not over-interpret the results obtained. After the aggregation experiment, different microscopies can be used, including the Atomic Force Microscopy (AFM) and the Transmission Electron Microscopy (TEM), in order to probe the morphologies of the aggregates formed.

Figure I-11. Representation of the Thioflavin T dye or ThT, with the free rotation around the bond between the two aromatic rings.

I-C Current treatments and therapeutic approaches

Nowadays, there is no known cure for AD. It exists only symptomatic treatments, improving the way of life of the patients. There are few drugs available. One of the categories is acetycholine esterase inhibitors. Indeed, acetylcholine (ACh) is a neurotransmitter involved in the memory and in the learning and it is hydrolysed by the acetylcholinesterase (AChE). ACh is deficient in the AD brains.

~ 52 ~

Chapter I: Context of the project

Inhibiting the action of the AChE enzyme should increase the concentration of ACh. One of the first AChE inhibitors was Tacrine (Figure I-12) which is rarely used today due to its side effects.116 Donepezil, Rivastigmine and Galantamine (Figure I-12) are the three AChE inhibitors currently available.116-118 The other group of symptomatic treatment is the antagonist of the N-methyl-D-aspartate (NMDA) receptor, involved for example in the memory. When NMDA receptors are activated by glutamate for example, NMDA receptors open their channel and Ca2+ can enter into the neuron. Nevertheless, in AD brains, there is an excitotoxity, meaning that NMDA receptors are over activated and the concentration of Ca2+ that enters into the neuron is too high, leading to the degradation of the neuron.119 The antagonists of the NMDA receptors block the channel and reduce the concentration of Ca2+ entering into the neuron.119 The actual available drug is the Memantine (Figure I-12).116-120 Some studies also propose high doses of vitamin E which is an antioxidant capable of improving the cognitive impairments,116 but clinical trials showed no benefit.

Figure I-12. The different structures of the available symptomatic drugs against AD.

Many other therapeutic approaches are currently under investigation.121-122 Some of them concern the degradation and clearance of Aβ aggregates.102 The immunotherapy is also highly under study.107, 123-124 The use of antibodies targeting Aβ peptide can allow the clearance of Aβ by the immune system. The most notable immunotherapies are Bapineuzumab which is a humanized monoclonal antibody and Solanezumab which is a monoclonal antibody.123 The first one has a high affinity for the 5 first amino acids of Aβ, and preferentially when Aβ is aggregated; the second one targets the center

~ 53 ~

Chapter I: Context of the project of Aβ (amino acids from 13 to 28) preferentially when Aβ is monomeric. Both of them have failed in clinical trials. Another human monoclonal antibody for the clearance of Aβ which is under clinical trials and show promising results is the Aducanumab.125 Another example of immunotherapy is the use of Interleukins 2 (IL2).126-127 Their concentration in the hippocampus in AD brains is lower than in healthy brains. Treatment with IL2 in AD mice shows improvement in the memory.126 This result is encouraging for the next immunotherapy investigations.

Some studies are developing small compounds to prevent the Aβ aggregation.107 Other ones focus on the inhibition of the Aβ production via the stimulation of the α-secretase,122 the modulation or inhibition of the γ-secretase in order to inhibit Aβ production or to induce the production of smaller Aβ peptides, not able to aggregate.107, 122, 128 The inhibition of β-secretase is also a way to prevent the Aβ production.107, 128 Furthermore, since the dyshomeostasis of Cu ions has been considered as a key event in the AD aetiology, the metal chelation approach is highly studied.116, 129-132 Multi-target compounds are also developed.122, 129, 131 The chelation moiety is linked to a moiety able to cross the BBB, or able to target Aβ aggregates, etc. It exists also therapeutic approaches concerning Tau protein, such as the inhibition of the aggregation of the protein.121-122 Another therapeutic approach focuses on the APOE ε4 allele on the apolipoprotein E (see I-A.ii).13 For example, using small molecules interfering with the domain interactions in APOE ε4, this apolipoprotein can be functionally and structurally converted to the APOE ε2 or 3.13

Many therapeutic approaches are under investigations. In this thesis, we have focused on the chelatotherapy targeting selectively Cu ions, since Cu ions are redox active and might stabilize small aggregates. Note that as Zn ions are essential in the neurotransmission and as they should be less toxic than Cu ions, they are to be kept in the synaptic cleft.

I-D Objectives of the study

Within the AD framework, Cu ion bound to Aβ in the synaptic cleft is currently considered as one of the target of choice due to its high toxicity: it is able to alter the aggregation of the Aβ peptide, leading to smaller more toxic aggregates and it is also able to catalyse ROS production. Note that these Cu ions bound to Aβ are then called “toxic Cu”. Many Cu(II) ligands have been studied, some of them have been tested in clinical trials, but failed to go through Phase III.130, 133 One of the hypothesis explaining this failure could be the lack of Cu ion selectivity of the ligands since they are also (and mainly) able to coordinate Zn ions. Up to now, the criteria for a Cu chelator against AD are the Blood Brain Barrier (BBB) permeability, the higher affinity constant for Cu ions than the Aβ, the affinity constant for Cu ions not too high in order to not remove Cu ions from other metalloproteins, leading

~ 54 ~

Chapter I: Context of the project to a more important dyshomeostasis (for more details, see § II-A). This study focuses on the Cu ion chelation. It is divided into two main parts.

The first part focuses on different Cu ion chelation strategies. First, a state of the art regards the strategy of using ligands, chelators or metallophores, in order to remove Cu ions from the Aβ peptide. In this thesis, a chelator is a ligand able to remove the metal ion from Aβ, while a metallophore is a chelator that redistribute metal ions into the cells where they are deficient. Their impacts on the ROS production, on the aggregation as well as their cell toxicity are reviewed. Then, two new concepts are developed. The first one detailed is the impact of the kinetic of Cu(II) removal from the Aβ peptide. Indeed, in order to remove Cu ions from the Aβ peptide, Cu ligands need a higher affinity constant for Cu ions than the one of Aβ. Nevertheless, this is a pre-requisite and could not be enough if the Cu chelation is too slow. To the best of our knowledge, this is the first time that this kinetic issue is studied in the AD context. The other proof of concept sheds light on the use of a Cu(I) and Cu(II) chelator. Indeed, nowadays, the redox state of Cu ions in the synaptic cleft is not well defined. This means that we do not know which Cu(I) or Cu(II) chelator will be the most efficient in the removal of Cu ions from Aβ. Therefore, a chelator able to remove both Cu(I) and Cu(II) is investigated in this context.

The second part focuses on the impact of Zn ions in the Cu ion chelation. This is biologically relevant since the concentration of Zn ions in the synaptic cleft should be 10 to 100 times higher than the concentration of Cu ions.58-59 Due to this high concentration, the ligands could chelate first Zn ions, leading to an inhibition of the Cu chelation, where Cu ions are the toxic target. First, a state of the art details the different studies about the mutual interactions between Cu ions and Zn ions with the Aβ peptide, as well as in the presence of a Cu-chelator to remove Cu ions from Aβ. The coordination of these metal ions with the peptide, their affinity constants, the impact of this interaction on the ROS production and on the aggregation as well as the impact on the Cu chelation are reported. Then, the first new concept of this part is the thermodynamic issue on the Cu chelation regarding Zn ions. The impact of the selectivity for Cu ions over Zn ions of the ligand compared to the one of the Aβ peptide is described and proved experimentally. The second new concept is a “pull-push” effect, demonstrating that, with a special category of ligands, Zn ions can trigger the Cu ion removal from the Aβ peptide. In other words, the presence of Zn(II) pulls Cu ion out of the Aβ peptide and pushes it inside the ligand, Zn(II) being bound to Aβ.

In brief, the investigations reported in the manuscript show new concepts required to develop more efficient ligand in the removal of Cu ions from the Aβ peptide.

~ 55 ~

Chapter I: Context of the project

References

1. Stelzma, R. A.; Schnitwlein, H. N.; Muriiagh, F. R., An English I’ranslation of Alzheimer’s 1907 Paper, “ijber eine eigenartige Erlranliung der Hirnrinde”. Clinical Anatomy 1995, 8, 429-431. 2. Ising, C.; Stanley, M.; Holtzman, D. M., Current Thinking on the Mechanistic Basis of Alzheimer’s and Implications for Drug Development. Clin. Pharmacol. Ther. 2015, 98 (5), 469-471. 3. AlzheimerAssociation, Basics of Alzheimer's disease 2016. 4. Qiu, C.; Kivipelto, M.; von Strauss, E., Epidemiology of Alzheimer's disease: occurrence, determinants, and strategies toward intervention. Dialogues Clin. Neurosci. 2009, 11 (2), 111-128. 5. 2017 alzheimer’s disease facts and figures. alz.org 2017. 6. Mayeux, R.; Stern, Y., Epidemiology of Alzheimer Disease. Cold Spring Harb. Perspect. Med. 2012, 2:a006239. 7. Janicki, S. C.; Schupf, N., Hormonal Influences on Cognition and Risk for Alzheimer’s Disease. Curr. Neurol. Neurosci. Rep. 2010, 10, 359-366. 8. Barnes, D. E.; Yaffe, K., The projected effect of risk factor reduction on Alzheimer's disease prevalence. Lancet Neurol. 2011, 10 (9), 819-828. 9. Norton, S.; Matthews, F. E.; Barnes, D. E.; Yaffe, K.; Brayne, C., Potential for primary prevention of Alzheimer's disease: an analysis of population-based data. Lancet Neurol. 2014, 13 (8), 788-794. 10. Daviglus, M. L.; Plassman, B. L.; Pirzada, A.; Bell, C. C.; Bowen, P. E.; Burke, J. R.; Connolly, E. S. J.; Dunbar-Jacob, J. M.; Granieri, E. C.; McGarry, K.; Patel, D.; Trevisan, M.; Williams, J. W. J., Risk factors and preventive interventions for Alzheimer disease: state of the science. Arch. Neurol. 2011, 68 (9), 1185-1190. 11. Letra, L.; Santana, I.; Seiça, R., Obesity as a risk factor for Alzheimer's disease: the role of adipocytokines. Metab. Brain Dis. 2014, 29 (3), 563-568. 12. Machado, A.; Herrera, A. J.; de Pablos, R. M.; Espinosa-Oliva, A. M.; Sarmiento, M.; Ayala, A.; Venero, J. L.; Santiago, M.; Villarán, R. F.; Delgado-Cortés, M. J.; Argüelles, S.; Cano, J., Chronic stress as a risk factor for Alzheimer's disease. Nat. Rev. Neurosci. 2014, 25 (6), 785-804. 13. Michaelson, D. M., APOE ε4: the most prevalent yet understudied risk factor for Alzheimer's disease. Alzheimers Dement. 2014, 10 (6), 861-868. 14. Imtiaz, B.; Tolppanen, A. M.; Kivipelto, M.; Soininen, H., Future directions in Alzheimer's disease from risk factors to prevention. Biochem. Pharmacol. 2014, 88 (4), 661-670. 15. Bilbul, M.; Schipper, H. M., Risk profiles of Alzheimer disease. Can. J. Neurol. Sci. 2011, 38 (4), 580-592. 16. Barber, R. C., The genetics of Alzheimer's disease. Scientifica (Cairo) 2012, 2012, 246210. 17. Cerajewska, T. L.; Davies, M.; West, N. X., Periodontitis: a potential risk factor for Alzheimer's disease. Br. Dent. J. 2015, 218 (1), 29-34. 18. Sivanandam, T. M.; Thakur, M. K., Traumatic brain injury: a risk factor for Alzheimer's disease. Neurosci. Biobehav. Rev. 2012, 36 (5), 1376-1381. 19. Schupf, N., Genetic and host factors for dementia in Down's syndrome. Br. J. Psychiatry. 2002, 180, 405-410. 20. Soto-Rojas, L. O.; de la Cruz-López, F.; Ontiveros Torres, M. A.; Viramontes-Pintos, A.; Cárdenas-Aguayo, M. C.; Meraz-Ríos, M. A.; Salinas-Lara, C.; Florán-Garduño, B.; Luna-Muñoz, J., Neuroinflammation and Alteration of the Blood-Brain Barrier in Alzheimer´s Disease. In Alzheimer's Disease - Challenges for the Future, Intech, Ed. 2015. 21. Mattson, M. P., Pathways towards and away from Alzheimer’s disease. Nature 2004, 430, 631- 639. 22. Jahn, H., Memory loss in Alzheimer’s disease. Dialogues Clin. Neurosci. 2013, 15 (4), 445-454. 23. Grundke-Iqbal, I.; Iqbal, K.; Tung, Y. C.; Quinlan, M.; Wisniewski, H. M.; Binder, L. I., Abnormal phosphorylation of the microtubule-associated protein tau (tau) in Alzheimer cytoskeletal pathology. PNAS 1986, 83, 4913-4917.

~ 56 ~

Chapter I: Context of the project

24. Silbert, L. C., Does statin use decrease the amount of Alzheimer disease pathology in the brain? Neurology 2007, 69, 8-11. 25. Holtzman, D. M.; Morris, J. C.; Goate, A. M., Alzheimer’s Disease: The Challenge of the Second Century. Sci. Transl. Med. 2011, 3 (77), 77sr1. 26. Laske, C.; Sohrabi, H. R.; Frost, S. M.; López-de-Ipiña, K.; Garrard, P.; Buscema, M.; Dauwels, J.; Soekadar, S. R.; Mueller, S.; Linnemann, C.; Bridenbaugh, S. A.; Kanagasingam, Y.; Martins, R. N.; O'Bryant, S. E., Innovative diagnostic tools for early detection of Alzheimer's disease. Alzheimers Dement. 2015, 11 (5), 561-578. 27. Weiner, M. W.; Veitch, D. P.; Aisen, P. S.; Beckett, L. A.; Cairns, N. J.; Green, R. C.; Harvey, D.; Jack, C. R. J.; Jagust, W.; Morris, J. C.; Petersen, R. C.; Saykin, A. J.; Shaw, L. M.; Toga, A. W.; Trojanowski, J. Q., Recent publications from the Alzheimer's Disease Neuroimaging Initiative: Reviewing progress toward improved AD clinical trials. Alzheimers Dement. 2017, 13 (4), 1-85. 28. Johnson, K. A.; Fox, N. C.; Sperling, R. A.; Klunk, W. E., Brain imaging in Alzheimer disease. Cold Spring Harb. Perspect. Med. 2012, 2 (4), a006213. 29. Caselli, R. J.; Beach, T. G.; Knopman, D. S.; Graff-Radford, N. R., Alzheimer Disease: Scientific Breakthroughs and Translational Challenges. Mayo Clin. Proc. 2017, 92 (6), 978-994. 30. Klunk, W. E.; Engler, H.; Nordberg, A.; Wang, Y.; Blomqvist, G.; Holt, D. P.; Bergström, M.; Savitcheva, I.; Huang, G. F.; Estrada, S.; Ausén, B.; Debnath, M. L.; Barletta, J.; Price, J. C.; Sandell, J.; Lopresti, B. J.; Wall, A.; Koivisto, P.; Antoni, G.; Mathis, C. A.; Långström, B., Imaging brain amyloid in Alzheimer's disease with Pittsburgh Compound-B. Ann. Neurol. 2004, 55 (3), 306-319. 31. Kantarci, K., Molecular imaging of Alzheimer disease pathology. Am. J. Neuroradiol. 2014, 35 (6), 12-17. 32. Rowe, C. C.; Ackerman, U.; Browne, W.; Mulligan, R.; Pike, K. L.; O'Keefe, G.; Tochon-Danguy, H.; Chan, G.; Berlangieri, S. U.; Jones, G.; Dickinson-Rowe, K. L.; Kung, H. P.; Zhang, W.; Kung, M. P.; Skovronsky, D.; Dyrks, T.; Holl, G.; Krause, S.; Friebe, M.; Lehman, L.; Lindemann, S.; Dinkelborg, L. M.; Masters, C. L.; Villemagne, V. L., Imaging of amyloid β in Alzheimer's disease with 18F-BAY94-9172, a novel PET tracer: proof of mechanism. Lancet Neurol. 2008, 7 (2), 129-135. 33. Wong, D. F.; Rosenberg, P. B.; Zhou, Y.; Kumar, A.; Raymont, V.; Ravert, H. T.; Dannals, R. F.; Nandi, A.; Brašić, J. R.; Ye, W.; Hilton, J.; Lyketsos, C.; Kung, H. F.; Joshi, A. D.; Skovronsky, D. M.; Pontecorvo, M. J., In Vivo Imaging of Amyloid Deposition in Alzheimer Disease Using the Radioligand 18F-AV-45 (Flobetapir F 18). J. Nucl. Med. 2010, 51 (6), 913-920. 34. Clark, C. M.; Schneider, J. A.; Bedell, B. J.; Beach, T. G.; Bilker, W. B.; Mintun, M. A.; Pontecorvo, M. J.; Hefti, F.; Carpenter, A. P.; Flitter, M. L.; Krautkramer, M. J.; Kung, H. F.; Coleman, R. E.; Doraiswamy, P. M.; Fleisher, A. S.; Sabbagh, M. N.; Sadowsky, C. H.; Reiman, E. M.; Zehntner, S. P.; Skovronsky, D. M., Use of florbetapir-pet for imaging β-amyloid pathology. JAMA 2011, 305 (3), 275- 283. 35. Nelissen, N.; Van Laere, K.; Thurfjell, L.; Owenius, R.; Vandenbulcke, M.; Koole, M.; Bormans, G.; Brooks, D. J.; Vandenberghe, R., Phase 1 Study of the Pittsburgh Compound B Derivative 18F- Flutemetamol in Healthy Volunteers and Patients with Probable Alzheimer Disease. J. Nucl. Med. 2009, 50 (8), 1251-1259. 36. Vandenberghe, R.; Laere, K. V.; Ivanoiu, A.; Salmon, E.; Bastin, C.; Triau, E.; Hasselbalch, S.; Law, I.; Andersen, A.; Korner, A.; Minthon, L.; Garraux, G.; Nelissen, N.; Bormans, G.; Buckley, C.; Owenius, R.; Thurfjell, L.; Farrar, G.; Brooks, D. J., 18F-flutemetamol amyloid imaging in Alzheimer disease and mild cognitive impairment: A phase 2 trial. Ann. Neurol. 2010, 68 (3), 319-329. 37. Bandara, N.; Sharma, A. K.; Krieger, S.; Schultz, J. W.; Han, B. H.; Rogers, B. E.; Mirica, L. M., Evaluation of 64Cu-Based Radiopharmaceuticals that Target Abeta Peptide Aggregates as Diagnostic Tools for Alzheimer's Disease. J. Am. Chem. Soc. 2017, 139 (36), 12550-12558. 38. Fodero-Tavoletti, M. T.; Villemagne, V. L.; Paterson, B. M.; White, A. R.; Li, Q.-X.; Camakaris, J.; O'Keefe, G. J.; Cappai, R.; Barnham, K. J.; Donnelly, P. S., Bis(thiosemicarbazonato) Cu-64 complexes for positron emission tomography imaging of Alzheimer's disease. J. Alzheimers Dis. 2010, 20 (1), 49- 55.

~ 57 ~

Chapter I: Context of the project

39. Torres, J. B.; Andreozzi, E. M.; Dunn, J. T.; Siddique, M.; Szanda, I.; Howlett, D. R.; Sunassee, K.; Blower, P. J., PET Imaging of Copper Trafficking in a Mouse Model of Alzheimer Disease. J. Nucl. Med. 2016, 57 (1), 109-114. 40. Jellinger, K. A., Alzheimer 100 – highlights in the history of Alzheimer research. J. Neural. Transm. 2006, 113, 1603–1623. 41. Olsson, B.; Lautner, R.; Andreasson, U.; Öhrfelt, A.; Portelius, E.; Bjerke, M.; Hölttä, M.; Rosén, C.; Olsson, C.; Strobel, G.; Wu, E.; Dakin, K.; Petzold, M.; Blennow, K.; Zetterberg, H., CSF and blood biomarkers for the diagnosis of Alzheimer's disease: a systematic review and meta-analysis. Lancet Neurol. 2016, 15 (7), 673-684. 42. Sperling, R.; Johnson, K., Biomarkers of Alzheimer disease: current and future applications to diagnostic criteria. Continuum (Minneap Minn) 2013, 19 (2), 325-338. 43. Chen, G.-f.; Xu, T.-h.; Yan, Y.; Zhou, Y.-r.; Jiang, Y.; Melcher, K.; Xu, H. E., Amyloid beta: structure, biology and structure-based therapeutic development. Acta Pharmacologica Sinica 2017, 1- 31. 44. O'Brien, R. J.; Wong, P. C., Amyloid Precursor Protein Processing and Alzheimer's Disease. Annu. Rev. Neurosci. 2011, 34 (1), 185-204. 45. Nicolas, M.; Hassan, B. A., Amyloid precursor protein and neural development. Development 2014, 141 (13), 2543-2548. 46. Sosa, L. J.; Caceres, A.; Dupraz, S.; Oksdath, M.; Quiroga, S.; Lorenzo, A., The physiological role of the Amyloid Precursor Protein (APP) as an adhesion molecule in the developing nervous system. J. Neurochem. 2017, In press. 47. Borghesani, V.; Alies, B.; Hureau, C., Cu(II) binding to various forms of amyloid-β peptides. Are they friends or foes ? Eur. J. Inorg. Chem. 2017, Accepted Manuscript. 48. Solomon, E. I.; Heppner, D. E.; Johnston, E. M.; Ginsbach, J. W.; Cirera, J.; Qayyum, M.; Kieber- Emmons, M. T.; Kjaergaard, C. H.; Hadt, R. G.; Tian, L., Copper Active Sites in Biology. Chem. Rev. 2014, 114 (7), 3659-3853. 49. Messerschmidt, A., Copper Metalloenzymes. In Comprehensive Natural Products II, Elsevier: Oxford, 2010; pp 489-545. 50. Elwell, C. E.; Gagnon, N. L.; Neisen, B. D.; Dhar, D.; Spaeth, A. D.; Yee, G. M.; Tolman, W. B., Copper–Oxygen Complexes Revisited: Structures, Spectroscopy, and Reactivity. Chem. Rev. 2017, 117 (3), 2059-2107. 51. Kozlowski, H.; Janicka-Klos, A.; Brasun, J.; Gaggelli, E.; Valensin, D.; Valensin, G., Copper, iron, and zinc ions homeostasis and their role in neurodegenerative disorders (metal uptake, transport, distribution and regulation). Coord. Chem. Rev. 2009, 253 (21), 2665-2685. 52. Eisses, J. F.; Kaplan, J. H., Molecular characterization of hCTR1, the human copper uptake protein. J. Biol. Chem. 2002, 277 (32), 29162-29171. 53. D'Ambrosi, N.; Rossi, L., Copper at synapse: Release, binding and modulation of neurotransmission. Neurochem. Int. 2015, 90, 36-45. 54. Scheiber, I. F.; Mercer, J. F.; Dringen, R., Metabolism and functions of copper in brain. Prog. Neurobiol. 2014, 116, 33-57. 55. Hung, Y. H.; Bush, A. I.; Cherny, R. A., Copper in the brain and Alzheimer’s disease. J. Biol. Inorg. Chem. 2010, 15, 61–76. 56. Ayton, S.; Lei, P.; Bush, A. I., Metallostasis in Alzheimer’s disease. Free Radical Biol. Med. 2013, 62, 76-89. 57. Donnelly, P. S.; Xiao, Z.; Wedd, A. G., Copper and Alzheimer's disease. Curr. Opin. Chem. Biol. 2007, 11 (2), 128-133. 58. Hureau, C., Coordination of redox active metal ions to the amyloid precursor protein and to amyloid-β peptides involved in Alzheimer disease. Part 1: An overview. Coord. Chem. Rev. 2012, 256 (19), 2164-2174. 59. Barnham, K. J.; Bush, A. I., Metals in Alzheimer's and Parkinson's Diseases. Curr. Opin. Chem. Biol. 2008, 12 (2), 222-228.

~ 58 ~

Chapter I: Context of the project

60. Kozlowski, H.; Luczkowski, M.; Remelli, M.; Valensin, D., Copper, zinc and iron in neurodegenerative diseases (Alzheimer’s, Parkinson’s and prion diseases). Coord. Chem. Rev. 2012, 256, 2129– 2141. 61. Hozumi, I.; Hasegawa, T.; Honda, A.; Ozawa, K.; Hayashi, Y.; Hashimoto, K.; Yamada, M.; Koumura, A.; Sakurai, T.; Kimura, A.; Tanaka, Y.; Satoh, M.; Inuzuka, T., Patterns of levels of biological metals in CSF differ among neurodegenerative diseases. J. Neurol. Sci. 2011, 303 (1), 95-99. 62. Barnham, K. J.; Bush, A. I., Biological metals and metal-targeting compounds in major neurodegenerative diseases. Chem. Soc. Rev. 2014, 43 (19), 6727-6749. 63. Frederickson, C. J.; Suh, S. W.; Silva, D.; Frederickson, C. J.; Thompson, R. B., Importance of Zinc in the Central Nervous System: The Zinc-Containing Neuron. J. Nutr. 2000, 130 (5), 1471S-1483S. 64. Kambe, T.; Tsuji, T.; Hashimoto, A.; Itsumura, N., The Physiological, Biochemical, and Molecular Roles of Zinc Transporters in Zinc Homeostasis and Metabolism. Physiol. Rev. 2015, 95 (3), 749-784. 65. Lovell, M. A.; Robertson, J. D.; Teesdale, W. J.; Campbell, J. L.; Markesbery, W. R., Copper, iron and zinc in Alzheimer’s disease senile plaques. J. Neurol. Sci. 1998, 158, 47–52. 66. Hureau, C.; Dorlet, P., Coordination of redox active metal ions to the amyloid precursor protein and to amyloid-β peptides involved in Alzheimer disease. Part 2: Dependence of Cu(II) binding sites with Aβ sequences. Coord. Chem. Rev. 2012, 256 (19), 2175-2187. 67. Migliorini, C.; Porciatti, E.; Luczkowski, M.; Valensin, D., Structural characterization of Cu2+, Ni2+ and Zn2+ binding sites of model peptides associated with neurodegenerative diseases. Coord. Chem. Rev. 2012, 256 (1), 352-368. 68. Rowinska-Zyrek, M.; Salerno, M.; Kozlowski, H., Neurodegenerative diseases – Understanding their molecular bases and progress in the development of potential treatments. Coord. Chem. Rev. 2015, 284, 298-312. 69. Kim, D.; Kim, N. H.; Kim, S. H., 34 GHz Pulsed ENDOR Characterization of the Copper Coordination of an Amyloid β Peptide Relevant to Alzheimer’s Disease. Angew. Chem. Int. Ed. 2013, 52 (4), 1139-1142. 70. Alies, B.; Renaglia, E.; Rózga, M.; Bal, W.; Faller, P.; Hureau, C., Cu(II) Affinity for the Alzheimer’s Peptide: Tyrosine Fluorescence Studies Revisited. Anal. Chem. 2013, 85 (3), 1501-1508. 71. Kowalik-Jankowska, T.; Ruta, M.; Wiśniewska, K.; Łankiewicz, L., Coordination abilities of the 1–16 and 1–28 fragments of β-amyloid peptide towards copper(II) ions: a combined potentiometric and spectroscopic study. J. Inorg. Biochem. 2003, 95 (4), 270-282. 72. Young, T. R.; Kirchner, A.; Wedd, A. G.; Xiao, Z., An integrated study of the affinities of the Aβ16 peptide for Cu(i) and Cu(ii): implications for the catalytic production of reactive oxygen species. Metallomics 2014, 6 (3), 505-517. 73. Atrián-Blasco, E.; Conte-Daban, A.; Hureau, C., Mutual interference of Cu and Zn ions in Alzheimer’s disease: perspectives at the molecular level. Dalton Trans. 2017, 46 (38), 12735–13146. 74. Conte-Daban, A.; Borghesani, V.; Sayen, S.; Guillon, E.; Journaux, Y.; Gontard, G.; Lisnard, L.; Hureau, C., Link between Affinity and Cu(II) Binding Sites to Amyloid-beta Peptides Evaluated by a New Water-Soluble UV-Visible Ratiometric Dye with a Moderate Cu(II) Affinity. Anal. Chem. 2017, 89 (3), 2155-2162. 75. Hureau, C.; Balland, V.; Coppel, Y.; Solari, P. L.; Fonda, E.; Faller, P., Importance of dynamical processes in the coordination chemistry and redox conversion of copper amyloid-β complexes. J. Biol. Inorg. Chem. 2009, 14 (7), 995-1000. 76. Shearer, J.; Szalai, V. A., The Amyloid-β Peptide of Alzheimer’s Disease Binds CuI in a Linear Bis- His Coordination Environment: Insight into a Possible Neuroprotective Mechanism for the Amyloid-β Peptide. J. Am. Chem. Soc. 2008, 130 (52), 17826-17835. 77. Himes, R. A.; Park, G. Y.; Siluvai, G. S.; Blackburn, N. J.; Karlin, K. D., Structural Studies of Copper(I) Complexes of Amyloid-β Peptide Fragments: Formation of Two-Coordinate Bis(histidine) Complexes. Angew. Chem. Int. Ed. 2008, 47 (47), 9084-9087. 78. Himes, R. A.; Park, G. Y.; Barry, A. N.; Blackburn, N. J.; Karlin, K. D., Synthesis and X-ray Absorption Spectroscopy Structural Studies of Cu(I) Complexes of HistidylHistidine Peptides: The Predominance of Linear 2-Coordinate Geometry. J. Am. Chem. Soc. 2007, 129 (17), 5352-5353.

~ 59 ~

Chapter I: Context of the project

79. Lu, Y.; Prudent, M.; Qiao, L.; Mendez, M. A.; Girault, H. H., Copper(i) and copper(ii) binding to β-amyloid 16 (Aβ16) studied by electrospray ionization mass spectrometry. Metallomics 2010, 2 (7), 474-479. 80. Alies, B.; Badei, B.; Faller, P.; Hureau, C., Reevaluation of Copper(I) Affinity for Amyloid-β Peptides by Competition with Ferrozine—An Unusual Copper(I) Indicator. Chem. Eur. J. 2012, 18 (4), 1161-1167. 81. Xiao, Z.; Gottschlich, L.; van der Meulen, R.; Udagedara, S. R.; Wedd, A. G., Evaluation of quantitative probes for weaker Cu(i) binding sites completes a set of four capable of detecting Cu(i) affinities from nanomolar to attomolar. Metallomics 2013, 5 (5), 501-513. 82. Danielsson, J.; Pierattelli, R.; Banci, L.; Graslund, A., High-resolution NMR studies of the zinc- binding site of the Alzheimer's amyloid beta-peptide. FEBS Journal 2007, 274 (1), 46-59. 83. Rezaei-Ghaleh, N.; Giller, K.; Becker, S.; Zweckstetter, M., Effect of Zinc Binding on β−Amyloid Structure and Dynamics: Implications for Aβ Aggregation. Biophys. J. 2011, 101 (5), 1202-1211. 84. Syme, C. D.; Viles, J. H., Solution 1H NMR investigation of Zn2+ and Cd2+ binding to amyloid-beta peptide (Aβ) of Alzheimer's disease. Biochimica et Biophysica Acta - Proteins and Proteomics 2006, 1764 (2), 246-256. 85. Alies, B.; Conte-Daban, A.; Sayen, S.; Collin, F.; Kieffer, I.; Guillon, E.; Faller, P.; Hureau, C., Zinc(II) Binding Site to the Amyloid-β Peptide: Insights from Spectroscopic Studies with a Wide Series of Modified Peptides. Inorg. Chem. 2016, 55 (20), 10499-10509. 86. Noel, S.; Bustos Rodriguez, S.; Sayen, S.; Guillon, E.; Faller, P.; Hureau, C., Use of a new water- soluble Zn sensor to determine Zn affinity for the amyloid-β peptide and relevant mutants. Metallomics 2014, 6 (7), 1220-1222. 87. Zawisza, I.; Rózga, M.; Bal, W., Affinity of copper and zinc ions to proteins and peptides related to neurodegenerative conditions (Aβ, APP, α-synuclein, PrP). Coord. Chem. Rev. 2012, 256 (19), 2297- 2307. 88. Chassaing, S.; Collin, F.; Dorlet, P.; Gout, J.; Hureau, C.; Faller, P., Copper and Heme-Mediated Abeta Toxicity: Redox Chemistry, Abeta Oxidations and Anti-ROS Compounds. Curr. Top. Med. Chem. 2012, 12 (22), 2573-2595. 89. Hureau, C.; Faller, P., Aβ-mediated ROS production by Cu ions: Structural insights, mechanisms and relevance to Alzheimer's disease. Biochimie 2009, 91 (10), 1212-1217. 90. Smith, D. G.; Cappai, R.; Barnham, K. J., The redox chemistry of the Alzheimer's disease amyloid β peptide. Biochim. Biophys. Acta 2007, 1768 (8), 1976-1990. 91. Barnham, K. J.; Masters, C. L.; Bush, A. I., Neurodegenerative diseases and oxidative stress. Nature Reviews 2004, 3, 205-214. 92. Reybier, K.; Ayala, S.; Alies, B.; Rodrigues, J. V.; Bustos Rodriguez, S.; La Penna, G.; Collin, F.; Gomes, C. M.; Hureau, C.; Faller, P., Free Superoxide is an Intermediate in the Production of H2O2 by Copper(I)-Aβ Peptide and O2. Angew. Chem. Int. Ed. 2015, 55 (3), 1085-1089. 93. Butterfield, D. A.; Lauderback, C. M., Lipid peroxidation and protein oxidation in Alzheimer’s disease brain: potential causes and consequences involving amyloid β-peptide-associated free radical oxidative stress. Free Radical Biol. Med. 2002, 32 (11), 1050-1060. 94. Butterfield, D. A.; Reed, T.; Newman, S. F.; Sultana, R., Roles of amyloid β-peptide-associated oxidative stress and brain protein modifications in the pathogenesis of Alzheimer's disease and mild cognitive impairment. Free Radical Biol. Med. 2007, 43 (5), 658-677. 95. Markesbery, W. R., Oxidative Stress Hypothesis in Alzheimer's Disease. Free Radical Biol. Med. 1997, 23 (1), 134-147. 96. Yatin, S. M.; Varadarajan, S.; Link, C. D.; Butterfield, D. A., In vitro and in vivo oxidative stress associated with Alzheimer’s amyloid β-peptide (1–42). Neurobiol. Aging 1999, 20 (3), 325-330. 97. Manevich, Y.; Held, K. D.; Biaglow, J. E., Coumarin-3-Carboxylic Acid as a Detector for Hydroxyl Radicals Generated Chemically and by Gamma Radiation. Radiat. Res. 1997, 148 (6), 580-591.

~ 60 ~

Chapter I: Context of the project

98. Valensin, D.; Gabbiani, C.; Messori, L., Metal compounds as inhibitors of β-amyloid aggregation. Perspectives for an innovative metallotherapeutics on Alzheimer's disease. Coord. Chem. Rev. 2012, 256 (19), 2357-2366. 99. Andreeva, T. V.; Lukiw, W. J.; Rogaev, E. I., Biological Basis for Amyloidogenesis in Alzheimer’s Disease. Biochemistry 2017, 82 (2), 122-139. 100. Minati, L.; Edginton, T.; Bruzzone, M. G.; Giaccone, G., Current Concepts in Alzheimer's Disease: A Multidisciplinary Review. Am J Alzheimers Dis Other Demen. 2009, 24 (2), 95-121. 101. Hardy, J. A.; Higgins, G. A., Alzheimer's disease: the amyloid cascade hypothesis. Science 1992, 256 (5054), 184-185. 102. Barage, S. H.; Sonawane, K. D., Amyloid cascade hypothesis: Pathogenesis and therapeutic strategies in Alzheimer's disease. Neuropeptides 2015, 52, 1-18. 103. Hardy, J.; Allsop, D., Amyloid deposition as the central event in the aetiology of Alzheimer's disease. Trends Pharmacol. Sci. 1991, 12 (10), 383-388. 104. Kozlov, S.; Afonin, A.; Evsyukov, I.; Bondarenko, A., Alzheimer’s disease: as it was in the beginning. In Nat. Rev. Neurosci., 2017. 105. Herrup, K., The case for rejecting the amyloid cascade hypothesis. Nature Neuroscience 2015, 18 (6), 794-799. 106. Armstrong, R. A., A critical analysis of the ‘amyloid cascade hypothesis’. Folia Neuropathol. 2014, 52 (3), 211-225. 107. Karran, E.; Strooper, B. D., The amyloid cascade hypothesis: are we poised for success or failure? J. Neurochem. 2016, 139 (S2), 237-252. 108. Faller, P.; Hureau, C.; Berthoumieu, O., Role of Metal Ions in the Self-assembly of the Alzheimer’s Amyloid-β Peptide. Inorg. Chem. 2013, 52 (21), 12193-12206. 109. Hellstrand, E.; Boland, B.; Walsh, D. M.; Linse, S., Amyloid beta-protein aggregation produces highly reproducible kinetic data and occurs by a two-phase process. ACS Chem. Neurosci. 2010, 1 (1), 13-18. 110. Viles, J. H., Metal ions and amyloid fiber formation in neurodegenerative diseases. Copper, zinc and iron in Alzheimer's, Parkinson's and prion diseases. Coord. Chem. Rev. 2012, 256 (19), 2271-2284. 111. Kayed, R.; Lasagna-Reeves, C. A., Molecular mechanisms of amyloid oligomers toxicity. J. Alzheimers Dis. 2013, 33, S67-S78. 112. Fu, L.; Sun, Y.; Guo, Y.; Chen, Y.; Yu, B.; Zhang, H.; Wu, J.; Yu, X.; Kong, W.; Wu, H., Comparison of neurotoxicity of different aggregated forms of Aβ40, Aβ42 and Aβ43 in cell cultures. J. Pept. Sci. 2017, 23 (3), 245-251. 113. Petkova, A. T.; Ishii, Y.; Balbach, J. J.; Antzutkin, O. N.; Leapman, R. D.; Delaglio, F.; Tycko, R., A structural model for Alzheimer's β-amyloid fibrils based on experimental constraints from solid state NMR. PNAS 2002, 99 (26), 16742-16747. 114. Tiiman, A.; Palumaa, P.; Tougu, V., The missing link in the amyloid cascade of Alzheimer’s disease – Metal ions. Neurochem. Int. 2013, 62, 367–378. 115. Khurana, R.; Coleman, C.; Ionescu-Zanetti, C.; Carter, S. A.; Krishna, V.; Grover, R. K.; Roy, R.; Singh, S., Mechanism of thioflavin T binding to amyloid fibrils. J. Struct. Biol. 2005, 151 (3), 229-238. 116. Rowinska-Zyrek, M.; Salerno, M.; Kozlowski, H., Neurodegenerative diseases – Understanding their molecular bases and progress in the development of potential treatments. Coord. Chem. Rev. 2015, 284, 298–312. 117. Santos, M. A.; Chand, K.; Chaves, S., Recent progress in repositioning Alzheimer’s disease drugs based on a multitarget strategy. Future Med. Chem. 2016, 8 (17), 2113-2142. 118. Tan, C. C.; Yu, J. T.; Wang, H. F.; Tan, M. S.; Meng, X. F.; Wang, C.; Jiang, T.; Zhu, X. C.; Tan, L., Efficacy and safety of donepezil, galantamine, rivastigmine, and memantine for the treatment of Alzheimer's disease: a systematic review and meta-analysis. J. Alzheimers Dis. 2014, 41 (2), 615-631. 119. Johnson, J. W.; Kotermanski, S. E., Mechanism of action of memantine. Curr. Opin. Pharmacol. 2006, 6 (1), 61-67. 120. Matsunaga, S.; Kishi, T.; Iwata, N., Memantine monotherapy for Alzheimer's disease: a systematic review and meta-analysis. PLoS One 2015, 10 (4), e0123289.

~ 61 ~

Chapter I: Context of the project

121. Hampel, H.; Schneider, L. S.; Giacobini, E.; Kivipelto, M.; Sindi, S.; Dubois, B.; Broich, K.; Nistico, R.; Aisen, P. S.; Lista, S., Advances in the therapy of Alzheimer's disease: targeting amyloid beta and tau and perspectives for the future. Expert Rev. Neurother. 2015, 15 (1), 83-105. 122. Berk, C.; Paul, G.; Sabbagh, M., Investigational drugs in Alzheimer's disease: current progress. Expert Opin. Investig. Drugs 2014, 23 (6), 837-846. 123. Tayeb, H. O.; Murray, E. D.; Price, B. H.; Tarazi, F. I., Bapineuzumab and solanezumab for Alzheimer's disease: is the ‘amyloid cascade hypothesis' still alive? Expert Opin. Biol. Ther. 2013, 13 (7), 1075-1084. 124. Morgan, D., Immunotherapy for Alzheimer’s Disease. J. Intern. Med. 2011, 269 (1), 54-63. 125. Sevigny, J.; Chiao, P.; Bussiere, T.; Weinreb, P. H.; Williams, L.; Maier, M.; Dunstan, R.; Salloway, S.; Chen, T.; Ling, Y.; O'Gorman, J.; Qian, F.; Arastu, M.; Li, M.; Chollate, S.; Brennan, M. S.; Quintero- Monzon, O.; Scannevin, R. H.; Arnold, H. M.; Engber, T.; Rhodes, K.; Ferrero, J.; Hang, Y.; Mikulskis, A.; Grimm, J.; Hock, C.; Nitsch, R. M.; Sandrock, A., The antibody aducanumab reduces Aβ plaques in Alzheimer's disease. Nature 2016, 537 (7618), 50-56. 126. Alves, S.; Churlaud, G.; Audrain, M.; Michaelsen-Preusse, K.; Fol, R.; Souchet, B.; Braudeau, J.; Korte, M.; Klatzmann, D.; Cartier, N., Interleukin-2 improves amyloid pathology, synaptic failure and memory in Alzheimer’s disease mice. Brain 2017, 140 (3), 826-842. 127. Dansokho, C.; Ait Ahmed, D.; Aid, S.; Toly-Ndour, C.; Chaigneau, T.; Calle, V.; Cagnard, N.; Holzenberger, M.; Piaggio, E.; Aucouturier, P.; Dorothée, G., Regulatory T cells delay disease progression in Alzheimer-like pathology. Brain 2016, 139 (4), 1237-1251. 128. Selkoe, D. J., Alzheimer’s Disease: Genes, Proteins, and Therapy. Physiol. Rev. 2001, 81 (2), 741–766. 129. Xia, N.; Liu, L., Metallothioneins and Synthetic Metal Chelators as Potential Therapeutic Agents for Removal of Aberrant Metal Ions from Metal-AB Species. Mini Rev. Med. Chem. 2014, 14 (3), 271- 281. 130. Robert, A.; Liu, Y.; Nguyen, M.; Meunier, B., Regulation of copper and iron homeostasis by metal chelators: a possible chemotherapy for Alzheimer's disease. Acc. Chem. Res. 2015, 48 (5), 1332- 1339. 131. Santos, M. A.; Chand, K.; Chaves, S., Recent progress in multifunctional metal chelators as potential drugs for Alzheimer’s disease. Coord. Chem. Rev. 2016, 327 –328, 287–303. 132. Scott, L. E.; Orvig, C., Medicinal inorganic chemistry approaches to passivation and removal of aberrant metal ions in disease. Chem. Rev. 2009, 109 (10), 4885-4910. 133. Sampson, E. L.; Jenagaratnam, L.; McShane, R., Metal protein attenuating compounds for the treatment of Alzheimer's dementia. Cochrane Database Syst. Rev. 2014, (2), Cd005380.

~ 62 ~

Chapter II: Cu ion chelation

Chapter II: Cu ion chelation

This chapter focuses on the Cu(I/II) chelation within the Alzheimer’s disease context. This work might be biologically relevant since Cu ions might interact with the Aβ peptide in AD as previously explained an form toxic Cu-Aβ species. Thus, Cu ions bound to Aβ were considered as a therapeutic target of choice.

A first part reports the state of art regarding the Cu(II) chelation approach in AD. Different ligands are grouped in categories, depending on their structures. Their impact on the ROS production by Cu- Aβ, on the aggregation of Aβ and on the cell toxicities are reported. Then, a first proof-of-concept focuses on the kinetic issue regarding the Cu(II) removal. Two series of macrocyclic ligands are studied. It is demonstrated that, by measuring the ROS production by Cu-Aβ, a high affinity constant for Cu(II) is not enough to remove the Cu ions from the Aβ if the removal is slower than the redox cycling of Cu- Aβ. The second proof-of-concept detailed here describes the removal of both Cu(II) and Cu(I) from the Aβ peptide since the redox state of Cu ions bound to Aβ in the synaptic cleft is not defined. The ligand L and the corresponding Cu(I) and Cu(II) complexes are characterized. The ability of this chelator to remove both Cu ions from the peptide is demonstrated by EPR and XANES spectroscopies as well as its ability to stop Cu-Aβ mediated ROS production.

II-A Cu ion chelators: State of art

As previously explained, the dyshomeostasis of Cu ions is considered as an important event in the AD aetiology. The metal chelation approach is intensively studied.1-5 One approach is preventing the coordination of Cu ions to Aβ using molecules able to occupy the binding site of Cu ions.6 Nevertheless, Cu ions have several binding sites so it is difficult to block all of them.6 Another approach is the use of ligands able to remove Cu ions from the Aβ peptide or to interfere with the Cu-Aβ complex in order to change its properties.3, 6-10 Up to now, different criteria for a Cu ligand to fight against AD exist. The ligand needs an affinity constant for Cu ions higher than the one of Aβ in order to remove the metal ion from the peptide. Nevertheless, this affinity constant does not have to be too much higher. Indeed, a too high affinity constant will lead to a ligand able to remove “healthy” Cu ions from essential metalloproteins.6, 11 This could exacerbate Cu ion dyshomeostasis in AD. Note that in this review, only ligands able to remove Cu ions from the Aβ1-x are reported. The N-terminally truncated peptides such as Aβ4-x and Aβ11-x are also present in AD brains and exhibit a higher affinity constant for Cu(II) than

~ 63 ~

Chapter II: Cu ion chelation the Aβ1-x (~ 1013).12-13 Hence, not all the ligands reported here are able to remove Cu(II) from the truncated peptides. Nevertheless, this should not be an issue since the Cu(II) complexes with the truncated peptides do not produce ROS.12-13

Another important criterion for the ligand in the AD context is its Blood Brain Barrier (BBB) permeability since the targeted Cu ions are localized in the brain.14 The Lipinski’s rules describes some criteria that have at least to be fulfilled for a passive diffusion: a low molecular weight, no more than 5 hydrogen bond donors, no more than 10 hydrogen bond acceptor and a log P < 5 (meaning a quite hydrophobic ligand).15 Regarding the hydrophobicity of the ligand, a balance between the hydrophilicity and the hydrophobicity is needed.2, 16-19 Ab initio studies propose a correlation between the metallo-aromaticity (i.e. “manifestation of aromatic properties in the chelate metallacycle”20) and the stability constant of the Cu(II)-ligand complex.21 Another approach to go through the BBB is the addition of a moiety to the ligand that increases its BBB permeability.14, 22 For example, ligands with a glucose moeity23 or grafted on a nanoparticle24-26 are developed.

Multi-target compounds are also developed in the AD chelatotherapy context.2, 4, 16, 19, 22, 27-29 Two strategies exist: the chelation moiety can be included in the multi-functionality or it can be linked to the multi-functionality moiety. The multi-target moiety can be an radical scavenging moiety, a moiety targeting the Aβ aggregates, a moiety able to cross the BBB as previously explained, etc. These compounds have many advantages. For example, they can target Aβ and the ligand can chelate Cu ions bound to Aβ and not to another metalloprotein. Another strategy is the use of prochelators. The chelating moiety is hidden; a cleavage of the molecule releases the chelating moiety. For example,

30 31 Franz’s group has proposed a cleavage by H2O2 or by β-secretase.

In this state of the art, only the Cu ion chelation is under focused: the multifunctional ligand will not be discussed, except if the chelator moiety is included inside the multi-functionality. First, the different structures of the ligands developed for the Cu chelatotherapy are described. The different ligands are gathered together in four categories: the hydroxyl/amino quinolones, the stilbene like and the benzothiazole (BzT) like ligands, the macrocyclic and peptidic ligands, and finally other structures. Then several aspects of the chelatotherapy in the AD context are discussed.

~ 64 ~

Chapter II: Cu ion chelation

II-A.i The different categories of ligands

Figure II-1. Three categories of ligands studied in the literature: the hydroxyl/aminoquinolines (top left), the benzothiazole like ligands (top right) and the stilbene like ligands (bottom).

Hydroxy-Aminoquinolines

The hydroxyl-aminoquinolines are one of the different categories of ligands developed against AD (Figure II-1, top left). The most well-known are clioquinol and then PBT2.6, 32 19, 33-40 Clioquinol (CQ) is a prototypical Metal Protein Attenuating Compound (MPAC). The MPAC are then called metallophores. Previously, it was tested against other diseases but was stopped due to its side effects. Nevertheless, it has been proposed and tested against AD.41 Many in vitro,42 such as structural characterizations of the Cu(II)-complex43-45 as well as aggregation studies46-47, studies on its capability to cross the BBB,48 and in vivo49-51 studies have been performed to determine its effects on the Aβ peptide and Cu ions. CQ was the first ligand against AD that has gone under clinical trials52-56 but has failed before going through phase III. Different hypotheses have been proposed, as for example the lack of selectivity (i.e. the ratio between the affinity constant for Cu(II) and the one for Zn(II)) between Cu and Zn ions. Indeed,

~ 65 ~

Chapter II: Cu ion chelation

CQ is able to chelate Cu and Zn ions with a very high affinity constant, meaning that in the synaptic cleft, it is able to chelate Zn(II) and to leave Cu ions bound to Aβ. Furthermore, it has also been stopped due to the issue in the scale-up of the fabrication. 57 Note that CQ is a metallophore, meaning that it is able to redistribute Cu ions inside the cells where they are deficient. Some studies on yeast for example have shown that CQ increases the intracellular concentration of Cu ions.58 The second generation of metallophores59 is the PBT2 synthetized by Barnham and collaborators.60 It is also a hydroxyquinoline, similar to CQ, but it does not have the iodine, supposed to trigger safety concerns.32 Some in vitro studies,61 studies on the solubility and permeability to BBB,57 in vivo tests and metallophoric capacities62 have been performed. PBT2 has also gone to clinical trials63-65 but failed in phase IIa. Then, other groups have developed hydroxyl/aminoquinolines based on the CQ moiety. The effect of these ligands are reminded in Table II-1. For the ROS production, two different effects are reported: redox silencing means that the ligand removes Cu ions from Aβ and stop the ROS production, radical scavenging means that the ligand itself captures the radical species. For the aggregation, the effect of the ligand by removing Cu ions from Aβ on the aggregation is reported as well as the effect of the ligand itself. Note that the modifications, such as linking two CQ moieties (see Figure II-1, top left), performed on the CQ moiety allow to have 1:1 complexes (ligand:metal ratio) and not 2:1 complexes as for CQ.

Table II-1. Hydroxy/aminoquinolines reported from the literature as well as their impact on the ROS production, on the aggregation and on cellular toxicities. P means that there is redox silencing or ROS scavenging, no Cu-induced aggregation or impact on the apo aggregation, no cell toxicity, metallophoric capabilities. O means that there is no redox silencing or no ROS scavenging, Cu-induced aggregation or no impact on the apo aggregation, cell toxicity, no metallophoric capabilities. WB stands for Western Blot, ThT means a ThT assay, TEM is the microscopy performed to image the aggregates. Note that if the characteristic has not been studied, n.d. (not determined) is written. pCu is calculated for a Cu(II) concentration at 10 µM. Grey lines mean that the complex formed is not a 1:1 complex. * = pH 7.1, ** = pH6.6, *** = pH 7.4. + = conditional value, ¤ = apparent value.

ROS production Aggregation Cell toxicity Radical Cu Wit- Metall- pCu Redox Ligand With Refs. Entry scave- removal Techn. hout ophore silencing effect Asc nging effect Asc Hydroxy/amino quinoline ThT + 46-47, 58, clioquinol 9.566 67 68 68 46-47 69 58 a P O ~ O AFM P P P 66-69 WB + ThT + PBT2 n.d. n.d. n.d. n.d. 71 61,70-71 b P O TEM61, P P 70 1 10.8¤*** P n.d. n.d. n.d. n.d. n.d. n.d. P 72-74 c PA1637 10.9¤*** P n.d. n.d. n.d. n.d. n.d. n.d. P 72, 75 d Micro 1 15.7*+ P n.d. P n.d. BCA n.d. n.d. n.d. 76 e Assay

~ 66 ~

Chapter II: Cu ion chelation

turb. + 8-H2QH n.d. n.d. P P O WB + P n.d. n.d. 70 f TEM turb. + 8-H2QS n.d. n.d. P P O WB + P n.d. n.d. 70 g TEM turb. + 8-H2QT n.d. n.d. P n.d. n.d. WB + n.d. n.d. n.d. 70 h TEM L 9.9*+ n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 59, 77 i ThT + 1 n.d. P n.d. O ~ WB + O ~ n.d. 67 j TEM

Stilbene like and benzothiazole like ligands

Stilbene and benzothiazole (BzT) like ligands form the same category due to their structures involved as Aβ targeting moieties (Figure II-1 top right and bottom). Indeed, BzT is the scaffold of the ThT and it is well known that ThT interacts with the β-sheets of the Aβ fibrils. Thus, ThT derivatives are developed as well as other planar structures such as the stilbene like. The effects of these ligands are reported in Table II-2. Regarding the ROS production, the impact of these ligands is not clear. Regarding the aggregation, BzT like ligands reported in Table II-2 do not have effect on the apo aggregation, apo meaning without any metal ions, while they have an effect on the metal-induced aggregation. Regarding the stilbene like ligands, some of them, meaning the ones with oxygen groups, have an impact on the apo-aggregation contrary to the other ones. The metal-induced aggregation is affected by these ligands, but again, the impact depends on the ligand. If cell toxicity has been studied, the results show that there is no cell toxicity for the ligand and/or the Cu(II)-complex in the presence or not of Aβ. No metallophore capability has been studied for these ligands.

Table II-2. Stilbene like and Benzothiazole like ligands reported from the literature as well as their impact on the ROS production, on the aggregation and on cellular toxicities. P means that there is redox silencing or ROS scavenging, no Cu- induced aggregation or impact on the apo aggregation, no cell toxicity, metallophoric capabilities. O means that there is no redox silencing or no ROS scavenging, Cu-induced aggregation or no impact on the apo aggregation, cell toxicity, no metallophoric capabilities. WB stands for Western Blot, ThT means a ThT assay, Turb. means a turbidimetric assay, IM-MS for ion mobility-mass spectrometry, TEM and AFM are the microscopies performed to image the aggregates. Note that if the characteristic has not been studied, n.d. (not determined) is written. Grey lines mean that the complex formed is not a 1:1 complex. pCu is calculated for a Cu(II) concentration at 10 µM. Grey lines mean that the complex formed is not a 1:1 complex. * = pH 7.1, ** = pH6.6, *** = pH 7.4. + = conditional value, ¤ = apparent value.

ROS production Aggregation Cell toxicity Radical Cu Wit- Metall- pCu Redox Ligand With Refs. Entry scave- removal Techn. hout ophore silencing effect Asc nging effect Asc Stilbene-like 2 n.d. ~ n.d. O n.d. ThT + P P n.d. 67 a

~ 67 ~

Chapter II: Cu ion chelation

WB + TEM ThT + 1 8.9+++¤ n.d. n.d. 68 b ~ P P P Turb. P ThT + 2 9.0+++¤ n.d. n.d. 68 c ~ P P P Turb. P ThT + 3 9.5+++¤ n.d. n.d. 68 d P O P P Turb. P ThT + 4 8.8+++¤ n.d. n.d. 68 e ~ P P P Turb. P ThT + 5 9.6+++¤ n.d. n.d. 68 f P ~ P P Turb. P ThT + 7a n.d. n.d. n.d. n.d. n.d. 61 g ~ P P TEM ThT + 8a n.d. n.d. n.d. n.d. n.d. 61 h ~ P ~ TEM ThT + 7b n.d. n.d. n.d. n.d. n.d. 61 i O P ~ TEM ThT + 8b n.d. n.d. n.d. n.d. n.d. 61 j O P ~ TEM WB + L2a n.d. n.d. n.d. n.d. 78 k ~ O O TEM P 1:1 : 7.8*+ WB + L2b n.d. n.d. n.d. 78-79 l 2:1 : P ~ O TEM P 6.8*+ WB + 1 n.d. n.d. n.d. 80 m ~ P O O TEM P WB + 2 n.d. n.d. n.d. 80 n ~ O P O TEM P WB + 3 n.d. n.d. n.d. 80 o ~ P O O TEM P WB + 4 n.d. n.d. n.d. n.d. 80 p P O ~ TEM P WB + ML 8.8*+ P P P P TEM / P n.d. n.d. 81 q IM-MS BzT-like Turb. + 1 6.0**+ n.d. n.d. 82 r ~ P O AFM P ~ Turb. + 2 n.d. n.d. n.d. 82 s ~ P O AFM P ~ Turb. + 3 n.d. n.d. n.d. 82 t ~ P O AFM P ~ 1:1 : 5.5*+ HBX n.d. n.d. Turb. n.d. n.d. n.d. 83 u 2:1 : ~ O 9.4*+ HBT 8.4*+ n.d. n.d. ~ O Turb. n.d. n.d. n.d. 83 v

~ 68 ~

Chapter II: Cu ion chelation

BM 8.1*+ n.d. n.d. ~ O Turb. n.d. n.d. n.d. 83 w

Figure II-2. Three other categories of ligands from the literature: salen, bispicen and bis(thiosemicarbazonato) (top left), macrocyclic ligands (top right) and peptidic ligands (bottom).

~ 69 ~

Chapter II: Cu ion chelation

Peptidic and macrocyclic ligands

Peptidic and macrocyclic ligands can belong to the same group (Figure II-2 bottom and top right). Indeed, the peptidic ligands studied here are rich in Histidine residues and they often can form ATCUN

84 complexes with Cu(II). The ATCUN motif is a NH2-Xxx-Yyy-His motif with a high affinity constant for Cu(II) (~ 1014) and with a high stability of the complex, precluding the switch between Cu(II) and Cu(I). The macrocyclic ligand form also Cu(II) complexes of high affinity, that is why peptidic and macrocyclic ligands can be grouped in the same category. The effects of these ligands are reported in Table II-3. This category of ligands does not have radical scavenging capability and no impact on apo aggregation. They are more focused on the removal of Cu ions from Aβ, leading to an important decrease of the ROS production catalysed by Cu-Aβ and to apo-like fibril formation.

Table II-3. Peptidic and macrocyclic ligands reported from the literature as well as their impact on the ROS production, on the aggregation. P means that there is redox silencing or ROS scavenging, no Cu-induced aggregation or impact on the apo aggregation, no cell toxicity, metallophoric capabilities. O means that there is no redox silencing or no ROS scavenging, Cu- induced aggregation or no impact on the apo aggregation, cell toxicity, no metallophoric capabilities. ThT means a ThT assay, Turb. means a turbidimetric assay, CD stands for circular dichroism, GFP stands for green fluorescent protein assay, TEM is the microscopy performed to image the aggregates. Note that if the characteristic has not been studied, n.d. (not determined) is written. pCu is calculated for a Cu(II) concentration at 10 µM. Grey lines mean that the complex formed is not a 1:1 complex. * = pH 7.1, ** = pH6.6, *** = pH 7.4. + = conditional value, ¤ = apparent value.

ROS production Aggregation Cell toxicity Radical Cu Wit- Metall- pCu Redox Ligand With Refs. Entry scave- removal Techn. hout ophore silencing effect Asc nging effect Asc Peptidic ligands (GH)2K 9.3*+ n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 85 a

(HH)2K 10.1*+ n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 85 b

KLVFF n.d. O ~ O O ThT n.d. O n.d. 86 c L 8.8*+ P n.d. n.d. n.d. n.d. n.d. n.d. n.d. 87 d AB12-20 n.d. P n.d. P P ThT n.d. ~ n.d. 86 e AB13-20 n.d. P n.d. P P ThT n.d. O n.d. 86 f GFP HWH 7.3***+ n.d. n.d. n.d. n.d. 88 g P P O ThT GFP HKcH 6.8***+ n.d. n.d. n.d. n.d. 88 h P P O ThT GFP HAH n.d. n.d. n.d. n.d. n.d. 88 i P P O ThT ~ GGH n.d.***¤ 4 eq n.d. n.d. n.d. n.d. n.d. n.d. 89 j P (4eq) 1 5.4*+ n.d. n.d. ~ O ThT n.d. n.d. n.d. 90 k

~ 70 ~

Chapter II: Cu ion chelation

2 5.3*+ n.d. n.d. ~ O ThT n.d. n.d. n.d. 90 l Macrocyclic ligands Turb. + cyclam 12.5*+ n.d. n.d. n.d. n.d. 91 m ~ P ThT P Turb. + cyclen 11.5*+ n.d. n.d. n.d. n.d. 91 n ~ P ThT P ThT + cyc-KLVFF n.d. n.d. n.d. n.d. n.d. 92 o P O TEM P cyc-LVFF n.d. P n.d. n.d. n.d. n.d. n.d. n.d. n.d. 92 p cyc- n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 92 q GKLVFF P Turb. + L1 (Yang) n.d. ~ n.d. P O CD + P ~ n.d. 93 r ThT Turb. + L2 (Yang) n.d. ~ n.d. P O CD + P ~ n.d. 93 s ThT

Other organic ligands

In this last category, salen, bispicen and bis(thiosemicarbazonato) ligands are grouped (Figure II- 2, top right). They mainly formed 1:1 complexes. The effects of these ligands are reported in Table II-4. They are able to remove Cu(II) (or Cu(I) for PTA ligand) from Aβ and to stop the associated deleterious events. The metallophoric capabilities have been studied only for the bis(thiosemicarbazonato) (btsc) ligands. 6,19, 34-35, 57, 94 Studies explain that some Cu(II)-btsc complexes, such as Cu(II)-gtsm, are able to cross the membrane of the cell and under the reductant intra-cellular conditions, Cu(II)-btsc is reduced. The “new” complex is not stable anymore and Cu(I) is released. Then, there is an activation of the Aβ degradation pathways.

Table II-4. Other ligands reported from the literature as well as their impact on the ROS production, on the aggregation. P means that there is redox silencing or ROS scavenging, no Cu-induced aggregation or impact on the apo aggregation, no cell toxicity, metallophoric capabilities. O means that there is no redox silencing or no ROS scavenging, Cu-induced aggregation or no impact on the apo aggregation, cell toxicity, no metallophoric capabilities. ThT means a ThT assay, Turb. means a turbidimetric assay, DLS stands for diffusion light scattering assay, TEM and AFM are the microscopies performed to image the aggregates Note that if the characteristic has not been studied, n.d. (not determined) is written. pCu is calculated for a Cu(II) concentration at 10 µM. Grey lines mean that the complex formed is not a 1:1 complex. * = pH 7.1, ** = pH6.6, *** = pH 7.4. + = conditional value, ¤ = apparent value.

ROS production Aggregation Cell toxicity Cu Radical Wit- Metall- pCu Redox remo- Ligand With Refs. Entry scave- Techn. hout ophore silencing val effect Asc nging Asc effect Other structures

*+ 95 H2GL1 8.6 n.d. P P n.d. Turb. n.d. n.d. n.d. a

~ 71 ~

Chapter II: Cu ion chelation

*+ 95 H2GL2 8.3 n.d. P ~ n.d. Turb. n.d. n.d. n.d. b ThT + L2 9.4*+ n.d. n.d. n.d. n.d. n.d. 96-97 c P P AFM ENDIP 10.2*+ n.d. n.d. P n.d. DLS n.d. n.d. n.d. 98 d DMAP 7.0*+ n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 98 e

ThT + PTA 13.2 ~ n.d. P20 n.d. n.d. n.d. n.d. 99 f equiv. AFM btsc n.d. n.d. n.d. n.d. n.d. n.d. P n.d. P 100-101 h

Discussion

All of the reported ligands have been separated in different categories, depending on their structures. Their effects on the ROS production, on the aggregation and on the cell toxicities have been reported. Note that the assays performed for each ligand are not the same. Regarding the aggregation assays, ThT assay, Western Blot or turbidimetry assay have been performed. The accuracy of the techniques are not the same and it becomes difficult to comprare the results. For example, with the turbidimetry assay, only the heterogeneity of the sample is probed and not the type of aggregates. For the ThT assay, some studies are performed during 24 h and other ones during one week. Another important issue when comparing the results is that the effect of the Cu removal on the ROS production as well as their radical scavenging capability depend on the ligand itself and also on the experimental condition (e.g. ratio ligand:metal:Aβ, order of addition of the reactants, etc.).

The first group, with the hydroxyl/aminoquinolines derivatives, the Bzt like and the stilbene like ligands, are quite planar ligands, supposed to interact with the β-sheets of the Aβ fibrils. They are supposed to impact the apo aggregation of Aβ. They are able to form 2:1 or 1:1 complexes (ligand:metal), depending on the ligand. This can lead to a difficulty in stopping the ROS production. Indeed, in a ternary complex, Cu ions could be in a not rigid geometry and the switch between Cu(II) and Cu(I) could be possible. Nevertheless, this group of ligands have most of the time a good radical scavenging capability (i.e. radical scavenging). Furthermore, for a metallophore, a 2:1 complex might not be a good point. Indeed, as the geometry of the complex is not rigid, it is possible to chelate both Cu(II) and Cu(I) easier than for a 1:1 complex. Thus, the release of Cu(I) upon reduction in the cell might be compromised by the high Cu(I) affinity constant.

The other important group is the macrocyclic and peptidic ligands and other ligands such as the salen. They exhibit a very high stability of their associated Cu(II) complex and they mainly form 1:1 Cu complexes. They are able to remove Cu(II) ions from Aβ and to stop the associated deleterious events

~ 72 ~

Chapter II: Cu ion chelation such as the ROS production and the Cu-induced aggregation. Nevertheless, they are not designed to have radical scavenging capabilities or interactions with the Aβ peptide. The bis(thiosemicarbazonato) ligands, mainly gtsm, show interesting metallophoric capabilities. They are able to redistribute Cu intra- cellularly, leading to the activation of the Aβ degradation pathways and improving the cognitive performances of AD mice.100-101

This section has reported the different ligands already studied in the context of the chelatotherapy against AD with their impacts on the ROS production, on the aggregation, on the cell toxicity and on their metallophoric capability.

II-B The kinetic aspect of the Cu(II) removal

This session focused on the importance of the kinetic of the Cu(II) chelation from Cu-Aβ in the AD context. It is composed of a draft of an article which will be submitted and of the supplementary information. The synthesis of the ligand, the potentiometric and crystallographic studies have been performed by the Tripier’s group in Brest. I have written the first draft.

We are currently waiting for EXAFS and EPR fitting data which should confirmed that the structures of the complexes studied are the same in solution than in crystals. Furthermore, the kinetic of the formation of the complexes will be performed later.

II-B.i Draft of the publication

Kinetic is the key when targeting copper ions to fight Alzheimer’s disease: an illustration with azamacrocyclic ligands

Amandine Conte-Daban,a,b Maryline Beyler,c Raphaël Tripier c and Christelle Hureau a,b

a) CNRS, LCC (Laboratoire de Chimie de Coordination), 205 route de Narbonne, BP 44099 31077 Toulouse Cedex 4, France b) University of Toulouse, UPS, INPT, 31077 Toulouse Cedex 4, France c) Université de Bretagne Occidentale, UMR-CNRS 6521, IBSAM, UFR des Sciences et Techniques, 6 avenue Victor le Gorgeu, C.S. 93837, 29238 BREST Cedex 3, France

~ 73 ~

Chapter II: Cu ion chelation

Abstract

Targeting copper ions to either remove or redistribute them is currently proposed as a possible therapeutic strategy in the context of Alzheimer’s disease (AD). Thermodynamic parameters, as for instance the Cu(II) affinity of the drug candidate or the Cu(II) over Zn(II) selectivity, are often/always considered. In contrast, kinetic ones have been overlooked despite their high importance. In the present communication, we use a wide series of azamacrocyclic ligands to demonstrate that kinetic issues has to be taken into account when designing copper chelators or metallophores in the context of AD.

AD is the most common neurodegenerative disorder characterized by the formation of extracellular senile plaques.102 They are detected in AD brains and contain high levels of Cu and Zn ions embedded with aggregates of the amyloid-β (Aβ) peptide.103-104 Aggregates of Aβ can be of various size and morphologies with distinct cell toxicity.105 While metal ions can effect aggregation pathways and final species, neither the mechanism nor the resulting aggregate types are clear-cut. Cu ions are essential and play key biological roles,106 but due to their redox ability, they play a harmful role in AD.107-108 Actually, Cu in the presence of a reductant such as ascorbate (Asc) present at 200-400 µM in the synaptic cleft, produces Reactive Oxygen Species (ROS) that participate in the oxidative stress observed in AD and damage neighboring biomolecules.108 This is one of the main reason why molecules targeting copper ions, mainly Cu(II), have been developed as drug candidates within the context of AD. Such ligands can either remove Cu from Aβ (chelator) and redistribute it (metallophore) and many of them have been reported in the last years.109-110 They fulfill several requirements, including having a higher affinity for Cu(II) than the Aβ peptide, the ability to redox silence the Cu(II) ion to stop the ROS production and a very high Cu over Zn selectivity.111 Tetraazamacrocycles are ligands well known for the kinetic inertness and the thermodynamic stability of the complexes counterparts. In particular, 1,4,8,11-tetrazacyclotetradecane (cyclam) and 1,4,7,10-tetraazacyclododecan (cyclen) have been previously tested in the AD context with mitigate results.112 They are also the scaffold of many chelators used for radiopharmaceutical purposes that bind 64/67Cu for positron emission tomography113-115 and Gd for magnetic resonance imaging116, for instance. As a general trend, Cu(II) complexation with unsubstituted tetraazamacrocycles is quite slow and pending arms are added to help improving the rate of Cu(II) complexation117-118 as it was observed also with other metals as Gd(III).119 Inspired by this observation, we report here the use of two families of tetraazamacrocycles based on the cyclen or cyclam scaffold as potential Cu(II) chelators against AD and illustrate the importance of kinetic issues in the removal of Cu(II) from the Aβ peptide.

~ 74 ~

Chapter II: Cu ion chelation

Scheme 1. Representation of the six different azaligands studied in this work.

Scheme 1 show the structure of the six ligands used in the present study. In addition to the non- functionalized cyclen and cyclam ligands, four macrocycles based on cyclen and cyclam ligands and bearing one (do1pa and te1pa, respectively) or two (do2pa and te2pa, respectively) methylpicolinate pendant arms are under study (Scheme S1). The last one is new and as for its three analogues, has been obtained following the regiospecific N-functionnalization of the macrocycle via the bisaminal chemistry (See synthetic scheme in the Supporting Information, Scheme S1). A view of the structure of the Cu(II) corresponding complex is shown in Figure 1, while bond distances and angles of the metal coordination environment are given in the caption (see also Tables S1 and S2). The central Cu(II) cation is six-coordinated in a highly elongated octahedral geometry, with the four nitrogen atoms of the macrocycle in the equatorial plane and the two nitrogen atoms of the picolinate arms at the apical positions. The Cu-NH bond are shorter than the Cu-Nalkylated bond ((1.9974(13) and 2.0462(13) Å, respectively) whereas the axial Cu-Nmethylpicolinate bonds are much longer (2.7910(16) Å). The carboxylate function does not participate to the coordination of the metal as it has already been observed with te1pa.120 This six-coordinated Cu(II) complex adopts the usually observed trans-III configuration (R,R,S,S conformation). The thermodynamic protonation constants of te2pa and its stability constants with Cu(II) were determined in 0.10 M KNO3 at 25°C using potentiometric titrations. The stepwise constants (log K) and the overall constants (log β) determined are collected in Table S3 while the speciation diagrams are presented in Figures S1-S2. Based on these values, the conditional affinity constant (corresponding to the absolute affinity constant at a given pH) has been calculated and equal 1014.9 at pH 7.1, a weaker value compared to affinity constants determined for the other two cyclam and te1pa analogues as well as for do2pa (Table 1).120-121 XANES and EPR signatures of the six Cu(II)-macrocylic complexes in frozen solution are shown in Figure S3-S4 and corresponding EPR parameters are listed in Table 1. XANES fingerprints of Cu(II)-te2pa is reminiscent of those of the parent Cu(II)-cyclam complex in line with the X-ray structure (Figure 1) that indicates an extremely long apical distance. EPR signatures and parameters are consistent with an elongated octahedral geometry and a 4N binding set in the equatorial plane.122

~ 75 ~

Chapter II: Cu ion chelation

Figure 1. X-ray crystal structures of Cu(II)-te2pa with atom labelling; hydrogen atoms and water molecules are omitted for simplicity. The ORTEP plots is drawn at the 30% probability level. Cu-N(1) = 2.0604(15), Cu-N(2) = 1.9906(15), Cu-N(3) = 2.7910(16), N(1)-Cu-N(2) = 86.61(6), N(1)-Cu-N(2) = 93.39(6), N(1)-Cu-N(3) = 75.81(5), N(2)-Cu-N(3) = 87.90(6), N(3)-Cu-N(1) = 104.19. Symmetry transformations used to generate equivalent atoms: #1 –x+1, -y+1, -z+1

The cyclic voltammogramm of the six Cu(II) complexes (Figure S5) indicate the possibility of reduction but at very low redox potentials (below -0.5 V versus SCE). In addition, the electrochemical reversibility of the cathodic process depend on the complex, with Cu(II)-do1pa, Cu(II)-do2pa and Cu(II)- te1pa showing quasi-reversible reduction process. As a general trend, addition of methylpicolinate arm(s) makes the reduction easier, i.e. the reduction occurs at higher potential (Table 1). Given the value of the cathodic peak, reduction by Asc is thermodynamically disfavored.

Table 1. Thermodynamic, spectroscopic and electrochemical characteristics of the six studied Cu(II) complexes (recorded under the very same conditions).

Cu(II)-cyclen Cu(II)-do1pa Cu(II)-do2pa Cu(II)-cyclam Cu(II)-te1pa Cu(II)-te2pa Coordination 4Neq 1Oap 5N 1O 5N 1O 4Neq 4 Neq 1Nap 4 Neq

pH7.1 17.4 – 18.8 18.5 17.4 19.9 – 20.8 18.0 14.9 refs. 123-124 120 121 125-126 120 This work

g// 2.20 ± 0.01 2.23 ± 0.01 2.23 ± 0.01 2.18 ± 0.01 2.19 ± 0.01 2.18 ± 0.01

65 -4 -1 A//( Cu) (10 cm ) 202 ± 5 187 ± 5 181 ± 5 224 ± 5 211 ± 5 217 ± 5

Epc (V vs. SCE) - 0.85 - 0.82 - 0.67 - 0.98 - 0.88 - 0.76 Reversibility Irrev. Rev. Rev. Irrev. Rev. Irrev. Incubation Cu P P P P P P Cu(Aβ) P P P P P P ROS No incubation Cu P P P O P P Cu(Aβ) O P P O P P

In line with the electrochemical measurements, once the Cu(II) complexes are formed (with any of the six ligands), they are not reduced by Asc as probed by the absence of Asc consumption

~ 76 ~

Chapter II: Cu ion chelation monitored by UV-Vis (Figure S6). Similarly, Asc consumption was not observed when Cu(II)-Aβ complex was pre-incubated with the ligand (Figure 2, panels A and B). This is in line with the higher Cu(II) affinity values of the ligands compared to Aβ (see Table 1) and the resistance to Asc reduction of the resulting Cu(II)-L complexes. Hence, the six macrocyclic ligands appear as good candidates for Cu(II) chelatotherapeutic purposes. In order to go further and challenge these ligands under more biologically relevant conditions, the Asc consumption assay was performed differently: Cu(II) (Figure S6) or Cu(II)-Aβ (Figure 2, Panels C and D) was first reacted with Asc and then the ligands were added. Such condition better mimics the brain environment, which is at the same time rich in Asc and dioxygen. Under such conditions, the arrest of Asc consumption depends on the ligand: the arrest is total with do1pa, te1pa and te2pa, quasi-total with do2pa and absent for both cyclam and cyclen. This may be linked to the rate of Cu(II) extraction from Cu(II)-Aβ during the redox cycling. Indeed, there is a competition between Cu(II)-Aβ reduction (and thus Asc consumption) and Cu(II) extraction by the ligands. If the Cu(II) extraction is not fast enough, the system Cu(II)-Aβ continue to redox cycle. Based on the properties of non-substituted and substituted ligands with respect to the rate of Cu(II) complexation, it is anticipated that the addition of the methylpicolinate arm(s) accelerates Cu(II) removal from Aβ.

Figure 2. Kinetics of Ascorbate consumption, followed by UV-visible spectroscopy at 265 nm with a background correction at 800 nm. Panel A. (a) Aβ16 + Cu(II) + Asc, (b) Aβ16 + Cu(II) + cyclen + Asc, (c) Aβ16 + Cu(II) + do1pa + Asc, (d) Aβ16 + Cu(II) + do2pa + Asc. Panel B. (a) Aβ16 + Cu(II) + Asc, (b) Aβ16 + Cu(II) + cyclam + Asc, (c) Aβ16 + Cu(II) + te1pa + Asc, (d) Aβ16 + Cu(II) + te2pa + Asc. Panel C. (a) Asc + Aβ16 + Cu(II), (b) Asc + Aβ16 + Cu(II) + cyclen, (c) Asc + Aβ16 + Cu(II) + do1pa, (d) Asc + Aβ16 + Cu(II) + do2pa. Panel D. (a) Asc + Aβ16 + Cu(II), (b) Asc + Aβ16 + Cu(II) + cyclam, (c) Asc + Aβ16 + Cu(II) + te1pa, (d) Asc + Aβ16 + Cu(II) + te2pa. [L] = [Aβ16] = 12 µM, [Cu(II)] = 10 µM, [Asc] = 100 µM, [HEPES] = 100 mM, pH 7.1. Note that the experiments are performed under aerobic conditions.

~ 77 ~

Chapter II: Cu ion chelation

It is worth noting the the Asc consumption experiments has previously been proved to mirror the

127 formation of ROS, in particular H2O2 and HO°. Here, we also directly monitor the formation HO° using the CCA Assay (Figure S7) and the results perfectly match those of the Asc consumption assay.

As modulation of Aβ aggregation by Cu ions could potentiate the toxicity of the Aβ aggregates, we also test the six ligands for the ability to prevent formation of Cu(II)-type fibrils and restore the formation of apo-like fibrils. All six ligands have the ability to do so, as evidenced either by the monitoring of kinetic of fibrils formation using the classical ThT assay and the morphologies of the Cu(II)-type or apo-type fibrils as imaged by TEM (Figures S8-S9). The identical activity of the six ligands observed here is linked to the time scale of the aggregation experiments (several hours), which allows Cu(II) to be removed from the Aβ peptide by any of them.

In the present communication, we report the study of six macrocyclic ligands to demonstrate how kinetic issues are important in the design of new chelator/metallophore within the AD context (Scheme 2). We in particular show that kinetic parameters can be as important as thermodynamic or structural one. This is particularly clear when comparing the te2pa and the parent cyclam ligand: While the Cu(II) environments are very close in both Cu(II) complexes and the affinity of the latter is higher by about 5 orders of magnitude, the te2pa can remove Cu(II) faster from Aβ and is active against ROS production in contrast to cyclam.

Scheme 2. Scheme representing the kinetic issues studied here, with the six macrocyclic ligands.

More generally, having ligands able to remove Cu from Aβ fast enough may be one new requirement to take into account in the design of drug-candidate against AD, which add to other

~ 78 ~

Chapter II: Cu ion chelation better-known requirement such as having an higher Cu affinity, being able to cross the blood brain barrier, not removing Cu ions form essential enzymes.

~ 79 ~

Chapter II: Cu ion chelation

II-B.ii Supplementary information

SUPPORTING INFORMATION

1-Synthesis of Cu(te2pa) complex

Scheme S1. (Top) Synthetic route to the te2pa ligand and its Cu(II) complex and (bottom) scheme of the six ligands under study.

~ 80 ~

Chapter II: Cu ion chelation

Compound 3.

Bisformyl-cyclam (1) (200 mg, 0.89 mmol) and NaI (540 mg, 3.57 mmol) were dissolved in 10 mL of freshly distilled acetonitrile. Methyl 6-(chloromethyl)picolinate (2) (348 mg, 1.88 mmol) dissolved in 10 mL of acetonitrile was added dropwise to the previous solution and the mixture was stirred at reflux for 5 days. The precipitate was filtered and washed with acetonitrile to obtain compound 3 as a white insoluble salt (686 mg, 98 %) that was used without further purification.

Ligand te2pa

An aqueous solution of NaOH 4 M was added to compound 3 (686 mg, 0.88 mmol). The mixture was stirred at r.t. for 3 days. HCl 6 M was added until pH 2. A white precipitate appeared that was filtered and washed with acetone to give te2pa as a hydrochloride salt (491 mg, 80 %).

1 3 4 H NMR (300 MHz, CDCl3): δ 8.05 (m, 4 H), 7.67 (dd, 2 H, J = 6.9 Hz, J = 2.1 Hz), 3.98 (s, 4H), 3.37 (bt, 4 H, 3J = 5.1 Hz), 3.02 (m, 4 H), 2.86 (bt, 4 H, 3J = 6.0 Hz), 1.97 (bq, 4 H, 3J = 6.3 Hz).

13 C NMR (75.47 MHz, CDCl3): 169.7, 160.0, 149.2, 130.6, 127.7, 60.1, 56.5, 52.1, 46.4, 46.2, 25.8

2+ + + HR-MS: m/z: 236.1396 [M + 2H] calcd. 236.1394 for C24H34N6O4+ 2H , 471.2718 [M+ H] calcd.

+ 471.2714 for C24H34N6O4 + H

Caution! Although no problem was found during our experiments, salts of perchlorate and their metal complexes are potentially explosive and should be handled with great care and in small quantities.

Cu(te2pa)

Te2pa.xHCl (50 mg, 0.081 mmol) was dissolved in H2O (5 mL) and the pH adjusted to 5 with a solution of KOH 1 M. Cu(ClO4)2.6H2O (39 mg, 0.11 mmol) was added to the ligand solution. The mixture was refluxed overnight and then concentrated. Acetone was added to precipitate the salts which were removed by filtration. This operation was repeated twice to lead to a purple solid (40.8 mg, 95 %).

2+ 2+ ESI-HR-MS (positive, H2O) m/z 266.5967[M + 2H] calcd. 266.5963 for [C24H32CuN6O4 + 2H] .

~ 81 ~

Chapter II: Cu ion chelation

2-X-ray diffraction study of Cu(II)-te2pa.10H2O

Table S1. Selected bond lengths (Å) and angles (°) of the metal coordination environment in Cu(II)-te2pa. See Figure 1 for labelling.

Cu(1)-N(1) 2.7910(16) Cu(1)-N(2) 1.9974(13) Cu(1)-N(3) 2.0462(13) N(2)-Cu(1)-N(2)#1 180.00(8) N(2)-Cu(1)-N(1) #1 92.95(6) N(2)-Cu(1)-N(1) 87.05(6)

Symmetry transformations used to generate equivalent atoms: #1 –x+1, -y+1, -z+1

Table S2. Crystal data and refinement details for Cu(II)-te2pa.10H2O

Cu(II)-te2pa.10H2O

formula C24H52CuN6O14 MW 712.26 crystal system Monoclinic

space group P21/c T/K 297(2) a/Å 9.5261(2) b/Å 12.2915(3) c/Å 14.3849(4) β/deg 101.102(2) V/Å3 1652.81(7) F(000) 758 Z 2

λ, Å (MoKα) 0.71073

-3 Dcalc/g cm 1.431 µ/mm-1 0.733 θ range/deg 3.26 to 30.50

~ 82 ~

Chapter II: Cu ion chelation

Rint 0.0232 reflns measd 16365 unique reflns 4955 GOF on F2 1.065

a R1 0.0551

b wR2 (all data) 0.1082 Largest differences peak and 0.456 and -0.283 hole /eÅ-3

a b 2 2 2 4 1/2 R1 = ∑Fo-Fc/ ∑Fo. wR2 = {∑[w(Fo -Fc ) ]/∑[ w(Fo )]}

3-Potentiometric study

Table S3. Stepwise and overall protonation and stability constants of te2pa2- and its Cu(II) complex at 25.0 ºC and I = 0.10 M in KNO3.

a Equilibrium reaction log βHiL / log KHiL L + H+ ⇄ HL 11.12 / 11.12 (2) + HL + H ⇄ H2L 21.48 / 10.36 (1)

+ H2L + H ⇄ H3L 24.86 / 3.38 (2)

+ H3L + H ⇄ H4L 27.49 / 2.63 (1)

log βMHiL / log KMHiL L + Cu(II) ⇄ CuL 23.5 / 23.5 (1) CuL + H+ ⇄ CuHL 26.47 /2.97 (5) CuL ⇄ CuLOH + H+ 12.47 / -10.93 (6) pCub 16.8 a L denotes the ligand in general; charges of ligand and complex species were omitted for simplicity, b 2+ -5 Calculated at pH = 7.4 for 100% excess of ligand with [M ]tot = 1.0×10 M

~ 83 ~

Chapter II: Cu ion chelation

100 LH2

LH 80 4

60 LH3 LH

40 L

% formation % relative toL 20

0 2 4 6 8 10 pH

-3 Figure S1. Speciation diagram of the protonated species of te2pa in aqueous solution at [L]tot = 10 M.

100 CuL CuLH 80

60

40 CuL(OH) CuLH-1

% formation% relativetoCu 20

0 2 4 6 8 10 pH

2+ -3 Figure S2. Speciation diagram of te2pa in presence of Cu(II) in aqueous solution at [M ]tot = [L]tot = 10 M.

~ 84 ~

Chapter II: Cu ion chelation

4-Spectroscopic signatures of the six Cu(II) complexes

Figure S3. XANES spectra of the different Cu(II) complexes. Panel A. (a) Cu(II)-cyclen, (b) Cu(II)-do1pa, (c) Cu(II)-do2pa. Panel B. (a) Cu(II)-cyclam, (b) Cu(II)-te1pa, (c) Cu(II)-te2pa. [L] = 1 mM, [Cu(II)] = 0.9 mM, [HEPES] = 100 mM, pH 7.1. 10% of glycerol was used as cryoprotectant. T = 20 K.

Figure S4. EPR signatures of the different Cu(II) complexes. Panel A. EPR experiments of (a) Cu(II)-cyclen, (b) Cu(II)-do1pa, (c) Cu(II)-do2pa. Panel B. EPR experiments of (a) Cu(II)-cyclam, (b) Cu(II)-te1pa, (c) Cu(II)-te2pa. [L] = 200 µM, [65Cu(II)] = 190 µM, [HEPES] = 50 mM, pH 7.1. 10% of glycerol was used as cryoprotectant. T = 110 K.

~ 85 ~

Chapter II: Cu ion chelation

5-Cyclic voltammetry signatures of the six Cu(II) complexes

Figure S5. Cyclic voltamograms of the different Cu(II) complexes. Panel A. Cu(II)-cyclen (orange curve), Cu(II)-do1pa (light green curve) and Cu(II)-do2pa (dark green line). Panel B. Cu(II)-cyclam (red curve), Cu(II)-te1pa (light blue curve) and Cu(II)- te2pa (dark blue line). [L] = 1.00 mM, [Cu(II)] = 0.96 mM, [phosphate buffer] = 100 mM at pH 7.1. The scanning speed is 0.1 V.s-1. Saturated Calomel Electrode was used as a reference.

~ 86 ~

Chapter II: Cu ion chelation

6-ROS production studies

Figure S6. Kinetics of Ascorbate consumption, followed by UV-visible spectroscopy at 265 nm with a background correction at 800 nm. Panel A. (a) Asc + Cu(II), (b) cyclen + Cu(II) + Asc, (c) do1pa + Cu(II) + Asc, (d) do2pa + Cu(II) + Asc. Panel B. (a) Asc + Cu(II), (b) cyclam + Cu(II) + Asc, (c) te1pa + Cu(II) + Asc, (d) te2pa + Cu(II) + Asc. Panel C. (a) Asc + Cu(II), (b) Asc + Cu(II) + cyclen, (c) Asc + Cu(II) + do1pa, (d) Asc + Cu(II) + do2pa. Panel D. (a) Asc + Cu(II), (b) Asc + Cu(II) + cyclam, (c) Asc + Cu(II) + te1pa, (d) Asc + Cu(II) + te2pa. [L] = [Aβ16] = 12 µM, [Cu(II)] = 10 µM, [Asc] = 100 µM, [HEPES] = 100 mM, pH 7.1. Asc is added latter after the first reactants in order to reach the thermodynamic equilibrium.

~ 87 ~

Chapter II: Cu ion chelation

Figure S7. Fluorescence kinetics of CCA experiments. Panel A. (a) Cu(II) + Asc, (b) cyclen-Cu(II) + Asc, (c) do1pa-Cu(II) + Asc, (d) do2pa-Cu(II) + Asc. Panel B. (a) Aβ16 + Cu(II) + Asc, (b) Aβ16 + Cu(II) + cyclen + Asc, (b’) Aβ16 + Cu(II) + Asc + cyclen, (c) Aβ16 + Cu(II) + do1pa + Asc, (c’) Aβ16 + Cu(II) + Asc + do1pa, (d) Aβ16 + Cu(II) + do2pa + Asc, (d’) Aβ16 + Cu(II) + Asc + do2pa. Panel C. (a) Cu(II) + Asc, (b) cyclam-Cu(II) + Asc, (c) te1pa-Cu(II) + Asc, (d) te2pa-Cu(II) + Asc. Panel D. (a) Aβ16 + Cu(II) + Asc, (b) Aβ16 + Cu(II) + cyclam + Asc, (b’) Aβ16 + Cu(II) + Asc + cyclam, (c) Aβ16 + Cu(II) + te1pa + Asc, (c’) Aβ16 + Cu(II) + Asc + te1pa, (d) Aβ16 + Cu(II) + te2pa + Asc, (d’) Aβ16 + Cu(II) + Asc + te2pa. If Asc is the last reactant, it was added 5 min after the beginning of the measurement. For the experiments A(b), A(c), A(d), B(b), B(c), B(d), C(b) and D(b), the samples were prepared 1 to 2 days before the addition of Asc. [L] = [Aβ16] = 12 µM, [Cu(II)] = 10 µM, [CCA] = 500 µM, [Asc] = 500 µM, [phosphate buffer] = 50 mM, pH 7.3.

~ 88 ~

Chapter II: Cu ion chelation

7-Aggregation study

Figure S8. Aggregation curves of Aβ peptide. Panel A. Kinetics of ThT fluorescence of Aβ (black curve), Cu(II)-Aβ (grey curve), Cu(II)-Aβ + cyclen (orange curve), Cu(II)-Aβ + do1pa (light green curve), Cu(II)-Aβ + do2pa (dark green curve). Panel B. Kinetics of ThT fluorescence of Aβ (black curve), Cu(II)-Aβ (grey curve), Cu(II)-Aβ + cyclam (red curve), Cu(II)-Aβ + te1pa (light blue curve), Cu(II)-Aβ + te2pa (dark blue curve). [L] = [Aβ40] = 20 µM, [Cu(II)] = 18 µM, [ThT] = 10 µM, [HEPES] = 50 mM, pH 7.1, T = 37°C.

Figure S9. TEM images of Aβ40 (left image), Cu(II)-Aβ (central image) and Aβ40 + Cu(II) + te2pa (right image).

8-Experimental details

Synthesis. Reagents were purchased from ACROS Organics and from ALDRICH Chemical Co. Cyclam was purchased from Chematech (Dijon, France). Bisformyl-Cyclam (1)128 and methyl 6- (chloromethyl)picolinate (2)129 were synthesized as previously described. The solvents were freshly distilled prior to use and according to the standard methods. NMR spectra (1H and 13C) were recorded at the core facilities of the University of Brest, with Bruker Avance 500 (500 MHz) or Bruker AMX-3 300 (300 MHz) spectrometers. The HR-MS analyses were performed at the Institute of Analytic and Organic Chemistry, ICOA in Orléans.

~ 89 ~

Chapter II: Cu ion chelation

Reagents, except the different ligands, were commercially available and were used as received. All the solutions were prepared in milliQ water (resistance: 18.2 MΩ.cm).

The Cu(II) ion source was CuSO4.5H2O, bought from Sigma-Aldrich. HEPES buffer (sodium salt of 2-[4-(2-hydroxyethyl)piperazin-1-yl]ethanesulfonic acid) was bought from Sigma-Aldrich. A stock solution was prepared at 500 mM, pH = 7.1.

Phosphate buffer was bought from Sigma-Aldrich. Two stock solutions, K2HPO4 and KH2PO4, were prepared at 500 mM, and they were mixed until to reach a stock solution at pH = 7.1. Sodium ascorbate was bought from Sigma-Aldrich. A stock solution was prepared at 5 mM each day because of the quick degradation of the ascorbate. Coumarin-3-carboxilic acid (CCA) was bought from Acros Organics. A stock solution at 5 mM was prepared in phosphate buffer at 500 mM, pH = 7.1. The stock solution was stored at 4°C. Thioflavin T (ThT) was bought from Acros Organics. A stock solution of ThT at 250 μM was prepared in water without any further purification.

Peptides. Aβ16 (DAEFRHDSGYEVHHQK) was bought from Genecust. A stock solution of about 10 mM was prepared and titrated using the Tyr chromophore, with ε = 1410 cm-1 M-1 at acidic pH. The stock solution was stored at 4°C. Aβ40 (DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVV) was bought from Genecust. Around 6 mg were dissolved in approx. 400 µL of NaOH 50 mM. This solution was purified by FPLC, with a Superdex 75 column and NaOH 15 mM as eluent, with a flow rate at 0.9 mL min-1. The collected fractions were titrated using the Tyr chromophore, with ε = 2400 cm-1 M-1 at basic pH. The stock solution was directly used for the ThT experiments.

Single crystal X-ray diffraction measurements. Single-crystal X-ray diffraction data were collected at 170 K on an X-CALIBUR-2 CCD 4-circle diffractometer (Oxford Diffraction) with graphite- monochromatized MoKα radiation (λ = 0.71073). Crystal data and structure refinement details are given in Table 5. Unit-cell determination and data reduction, including interframe scaling, Lorentz, polarization, empirical absorption and detector sensitivity corrections, were carried out using attached programs of Crysalis software (Oxford Diffraction).130 Structures were solved by direct methods and refined by full matrix least squares method on F2 with the SHELXL131 suites of programs. The hydrogen atoms were identified at the last step and refined under geometrical restraints and isotropic U- constraints.132 CCDC number 1540075-1540076 contains the supplementary crystallographic data for this paper. These data can be obtained free of charge from the Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.

~ 90 ~

Chapter II: Cu ion chelation

Potentiometric studies.

Equipment and Work Conditions The potentiometric setup consisted of a 50 mL glass-jacketed titration cell sealed from the atmosphere and connected to a separate glass-jacketed reference electrode cell by a Wilhelm type salt bridge filled with 0.1 M KNO3 electrolyte. An Orion 720A+ measuring instruments fitted with a Metrohm 6.0123.100 glass electrode and a Metrohm 6.0733.100 Ag/AgCl reference electrode was used for the measurements. Batch points were measured with a Metrohm 6.0233.100 combined glass electrode. The ionic strength of the experimental solutions was kept at 0.10 ± 0.01 M with KNO3; temperature was controlled at 298.2 ± 0.1 K using a Huber CC3-K6 compact cooling and heating bath thermostat and a previously calibrated Orion 91-70-06 ATC-probe.

Atmospheric CO2 was excluded from the titration cell during experiments by slightly bubbling purified nitrogen on the experimental solution. Titrant solutions were added through capillary tips at the surface of the experimental solution by a Metrohm Dosimat 665 automatic buret. Titration procedure is automatically controlled by software after selection of suitable parameters, allowing for long unattended experimental runs. The titrant was a KOH solution prepared at ca. 0.1 M from a commercial ampule of analytical grade, and its accurate concentration was obtained by application of

133 the Gran’s method upon titration of a standard HNO3 solution. Ligand solution was prepared at ca. 2.0 × 10–3 M, and the Cu(II) solution was prepared at ca. 0.05 M from analytical grade nitrate salts and

134 standardized by complexometric titrations with H4edta (ethylenediaminetetraacetic acid). Sample solutions for titration contained approximately 0.04 mmol of ligand in a volume of 30.00 mL. In complexation titrations metal cations were added at 0.9 equiv of the ligand amount. In competition titrations H4edta was additionally added at 1.2 equiv. Batch titrations were prepared in a similar way with approximately 0.08 mmol of the ligand in a total volume of 3.00 mL, with Cu(II) and H4edta added respectively at 1 and 2.3 equiv. of the ligand amount. Increasing amounts of standardized KOH solution at ca. 0.1 M were added to each one. Batch titration points were incubated in tightly closed vials at 25 °C until potential measurements attained complete stability.

Measurements. The electromotive force of the sample solutions was measured after calibration of the

–3 + electrode by titration of a standard HNO3 solution at 2.0 × 10 M in the work conditions. The [H ] of the solutions was determined by measurement of the electromotive force of the cell, E = E°′ + Q log

+ + [H ] + Ej. The term pH is defined as −log[H ]. E°′ and Q were determined by acid region of the calibration curves. The liquid-junction potential, Ej, was found to be negligible under the experimental conditions

+ – –13.78 used. The value of Kw = [H ][OH ] was found to be equal to 10 by titrating a solution of known hydrogen-ion concentration at the same ionic strength in the alkaline pH region, considering E°′ and Q valid for the entire pH range. The protonation constants of H4edta and the thermodynamic stability

~ 91 ~

Chapter II: Cu ion chelation constants of its copper(II) complex used in competition titration refinements were taken from the literature.135 Each titration consisted of 50–70 equilibrium points in the range pH 2.5-11.5, and at least two replicate titrations were performed for each particular system.

Calculations. The potentiometric data were refined with the HYPERQUAD software,136 and speciation

137 H diagrams were plotted using the HySS software. The overall equilibrium constants βi and βMmHhLl are

m h l H h l defined by βMmHhLl = [MmHhLl]/[M] [H] [L] (βi = [HhLl]/[H] [L] and βMH–1L= βML(OH) × Kw). Differences, in log units, between the values of protonated (or hydrolyzed) and nonprotonated constants provide the stepwise (log K) reaction constants (being KMmHhLl = [MmHhLl]/[MmHh–1Ll][H]). The errors quoted are the standard deviations calculated by the fitting program from all the experimental data for each system.

Electron Paramagnetic Resonance. Electron Paramagnetic Resonance (EPR) data were recorded using an Elexsys E 500 Bruker spectrometer, operating at a microwave frequency of approximately 9.5 GHz. Spectra were recorded using a microwave power of 5 mW across a sweep width of 120 mT (centred at 310 mT) with modulation amplitude of 1.0 mT. Experiments were carried out at 110 K using a liquid nitrogen cryostat.

EPR samples were prepared from stock solution of ligand diluted down to 0.2 mM in H2O. 0.95 eq. of

65 65 65 Cu(II) was added from 25 mM Cu(NO3)2 stock solution home-made from a Cu foil. If necessary, pH was adjusted with H2SO4 and NaOH solutions. Samples were frozen in quartz tube after addition of 10% glycerol as a cryoprotectant and stored in liquid nitrogen until used.

Electrochemistry. Cyclic voltamogram were recorded on an Autolab PGSTAT302N at 25°C. Saturated Calomel Electrode was used as a reference, Platine electrode was the counter electrode and the working electrode was a glassy carbon electrode. This last electrode was carefully polished before each measurement on a red disk NAP with 1 µm AP-A suspension under abundant distillate water flow during at least three minutes (Struers). The solution was deoxygenated by bubbling Argon before each measurement. Any support electrolyte was added because of the high concentration of phosphate buffer in the solution. The scanning speed was 0.1 V.s-1. The samples were prepared from stock solutions of ligand and Cu(II) down to approx. 1 mM and 0.9 mM respectively in a buffered solution.

UV-Visible spectrophotometry. UV-vis kinetics were recorded on a spectrophotometer Agilent 8453 at 25°C in 1 cm path length quartz cuvette, with an 800 rpm stirring. The samples were prepared from stock solutions of ligand, peptide and Cu(II) diluted down to 12, 12 and 10 µM respectively in HEPES solution, pH = 7.1. Ascorbate is diluted down to 100 µM.

~ 92 ~

Chapter II: Cu ion chelation

CCA experiments were recorded on a FLUOstar OPTIMA BMG LABTECH at 25°C in a 96-well plate bought from Dutscher SAS. CCA was excited at 390 nm and the fluorescence was recorded at 450 nm. The gain was 1350. The samples were prepared from stock solutions of ligand, peptide and Cu(II) diluted down to 12, 12 and 10 µM respectively in phosphate solution, pH = 7.1. CCA was added at a resulting concentration of 500 µM. Injector was used for the addition of ascorbate diluted down to 500 µM, 5 min after the beginning of the experiment.

ThT experiments were recorded on a FLUOstar OPTIMA BMG LABTECH at 37°C in a 384-well plate bought from Dutscher SAS. ThT was excited at 440 nm and the fluorescence was recorded at 490 nm. The gain was 1200. The samples were prepared from stock solutions of ligand, peptide and Cu(II) diluted down to 20, 20 and 18 µM respectively in HEPES buffer, pH = 7.1. ThT was added at a resulting concentration of 10 µM.

Transmission electron microscopy. Specimens were prepared for electron microscopy using the conventional negative staining procedure. 20 μL of solution was absorbed on Formvar-carbon-coated grids for 2 min, blotted, and negatively stained with uranyl acetate (1%) for 1 min. Grids were examined with a TEM (Jeol JEM-1400, JEOL Inc, Peabody, MA, USA) at 80 kV. Images were acquired using a digital camera (Gatan Orius, Gatan Inc, Pleasanton, CA, USA) at a x 25 000 magnification.

~ 93 ~

Chapter II: Cu ion chelation

II-C A Cu(I) and Cu(II) chelator

This session focused on a chelator able to chelate Cu(II) but also Cu(I). It is composed of a summary of the article published in Chemistry – A European Journal in 2017, then the publication itself and finally the supporting information. The synthesis of the ligand, the potentiometric studies and the mass spectroscopy analyses have been performed by the Delangle’s group in Grenoble.

II-C.i Summary

This article describes a chelator able to remove both Cu(II) and Cu(I) from the Aβ peptide within the context of Alzheimer’s disease. In most cases of the literature, only Cu(II) chelators are studied, although the redox state of Cu ions bound to Aβ in the synaptic clefts is not known. Moreover, as previously detailed, the Cu-Aβ complex has the capability to produce ROS in presence of a reductant and O2. This means that Cu ions cycle between the (+I) and the (+II) redox state during the ROS production. Mostly Cu(II) (or Cu(I)) chelators in the literature that are redox-silent have a coordination optimized for Cu(II) (or Cu(I)). Thus, Cu(II) ligands are classically poor Cu(I) ligands and vice versa. Thus, a ligand that binds Cu(II) and Cu(I) strongly is new and it is explained in the following why it is also redox-silent.

This ligand, noted L, is designed in order to chelate both redox states of Cu ions. As for the Aβ peptide, three Histidine residues are involved in the coordination of the metal ion. They are graft on a nitrilotriacetic platform. L has affinities higher than Aβ for Cu(II) and Cu(I) and should be able to remove Cu(II) and Cu(I) from the peptide.

The first step of this study is the synthesis of the ligand. Three Histidine residues are grafted on a nitrilotriacetic acid scaffold. The carboxylate groups of the Histidine residues are amidated to avoid their possible coordination to metal ions. Then, the protonation states are measured.

The next step was the characterization of the complex Cu(II)-L. ESI-MS has shown the existence of the complex 1:1 as the unique species. Then, potentiometric studies have been performed and demonstrated three predominant species between pH 3 and pH 8. This leads to a proposition of three binding modes depending on the pH. The first one, at pH around 4, involves two nitrogen atoms from Histidine residues, the third one staying protonated. Two molecules of solvent or O atoms from the CO groups complete the coordination sphere. Around pH 6, the predominant form involves the three Hisitine residues and a solvent molecule or an O atom. Finally, at pH around 7, the three Histidine residues and a deprotonated amide coordinate the Cu(II). The last characterization of the complex has

~ 94 ~

Chapter II: Cu ion chelation been performed by EPR. The deduced parameters confirm the three binding modes previously proposed at the different pH.

The third step of this study was the characterization of the Cu(I)-L complex. As for the Cu(II) complex, the ESI-MS has shown a 1:1 complex as the unique species. Then, EXAFS experiments have been performed. The intensity of the 1s ‰ 4p transition of Cu(I) corresponds to a tri- or tetra- coordination. The fitting of the EXAFS data proposes three nitrogens (likely from the Histidine residues) and one oxygen as ligands for the Cu(I). Then, the affinity constant of L for Cu(I) has been determined by a competition experiment, with ferrozine as a competitor. As expected, for Cu(II) and for Cu(I), L shows a higher affinity constant than Aβ, due to its pre-organized scaffold.

The last step of the characterization of the complexes was their electrochemical study. The Cu(I)/Cu(II) redox process is characteristic of a ECEC (E = electrochemical, C = chemical) electrochemical process. Cu(II), under a square planar geometry, is reduced into Cu(I) under the same geometry (E), which reorganizes into a tetrahedral geometry (C). For the reverse oxidation process, the oxidation occurs on Cu(I) in the tetrahedral geometry (E), and then the species rearranges into a square planar geometry (C). Another important parameter is that the reduction of Cu(II)-L occurs at a lower potential than the oxidation of the Ascorbate, and the oxidation of Cu(I)-L occurs at higher potential than the reduction of O2. This means that the complexes should resist to the reactions with ascorbate and O2, i.e. if L removes Cu ions from the peptide, the ROS production would be slowed down significantly.

The ability of L to remove Cu(II) and Cu(I) from the Aβ peptide has been studied by EPR and XANES spectroscopies, respectively. The EPR experiments have shown that one equivalent of L is enough to remove almost all Cu(II) from Aβ. A linear combination of the signatures of Cu(II)-Aβ and Cu(II)-L proposes that less than 5 % of Cu(II) stays on Aβ. Based on the affinity constants, 0.2 % of Cu(II) should stay on the peptide. The XANES experiments have probed the removal of Cu(I). A linear combination of the spectra of Cu(I)-Aβ and Cu(I)-L shows that around 17 % ± 5 % of Cu(I) stay bound to Aβ. This value is also in good agreement with the affinity constants since 30 % of Cu(I) should stay bound to the peptide. These experiments have proved that L is able to fully remove Cu(II) and partly Cu(I) from the Aβ peptide, in line with the affinity constants.

The last part of this work was the study of the capability of L to stop the ROS production catalyzed by Cu-Aβ. Two indirect methods have been used: the consumption of ascorbate by UV-Vis spectroscopy and the following of 7-OH-CCA formation by fluorescence. The UV-Vis experiments have shown that L stops the ROS production in all the configurations: if the Cu(II) or Cu(I) complex is formed before the addition of ascorbate, as well as if L is added during the ROS production. The results

~ 95 ~

Chapter II: Cu ion chelation obtained by CCA fluorescence assay are consistent. Note that this is not intuitive. Indeed, if a ligand chelates Cu(I) and Cu(II), it may be able to catalyze the ROS production. The hypothesis for L is that the geometry and coordination modes for the Cu(I) and Cu(II) complexes are very different. Switching from one complex to another one should take too much time, leading to a strong decrease of the ROS production.

Figure II-3. Scheme representing the ligand L, able to remove both Cu(I) and Cu(II) from Aβ and stop the ROS production.

In summary, this article reports a ligand able to chelate Cu(II) and Cu(I) (Figure II-3). Different characterizations have been performed for both complexes. The binding modes have been proposed, depending on the pH. Affinity constants have been measured. The electrochemical properties of the complexes have shown that after the reduction or oxidation of the metal, a geometric reorganization occurs. In addition, the ability of L to remove Cu(II) and Cu(I) from Aβ has been probed by EPR and XANES, and 5 % of Cu(II) and 17 % of Cu(I) stay bound to the peptide. Then the capability of L to stop the ROS produced by Cu-Aβ has been proved. The strategy of targeting both Cu(I) and Cu(II) ions shows interesting properties in the context of the removal of Cu ions in the Alzheimer’s disease. Furthermore, this could overcome the kinetic issue for a Cu(I) or Cu(II) ligand. Indeed, if the ligand takes more time to remove Cu ions from Aβ than the redox process of Cu-Aβ, the ligand will not be able to remove Cu from Aβ. If the ligand can chelate both metal ions, there is no more kinetic issue.

~ 96 ~

Chapter II: Cu ion chelation

II-C.ii Article

~ 97 ~

Chapter II: Cu ion chelation

~ 98 ~

Chapter II: Cu ion chelation

~ 99 ~

Chapter II: Cu ion chelation

~ 100 ~

Chapter II: Cu ion chelation

~ 101 ~

Chapter II: Cu ion chelation

~ 102 ~

Chapter II: Cu ion chelation

~ 103 ~

Chapter II: Cu ion chelation

~ 104 ~

Chapter II: Cu ion chelation

~ 105 ~

Chapter II: Cu ion chelation

~ 106 ~

Chapter II: Cu ion chelation

~ 107 ~

Chapter II: Cu ion chelation

II-C.iii Supporting information

A trishistidine pseudopeptide with ability to remove both Cu(I) and Cu(II) from the amyloid-β peptide and to stop the associated ROS formation

A. Conte-Daban,a,b§ B. Boff,c§ A. Candido Matias,c,d C. N. Montes Aparicio,a,b C. Gateau,c C. Lebrun,c G. Cerchiaro,d I. Kieffer,e,f S. Sayen,g E. Guillon,g P. Delangle,c* and C. Hureaua,b*

a. CNRS, LCC (Laboratoire de Chimie de Coordination), 205 route de Narbonne, BP 44099 31077 Toulouse Cedex 4, France. b. University of Toulouse, UPS, INPT, 31077 Toulouse Cedex 4, France. c. Univ. Grenoble Alpes, CEA, CNRS, INAC, SyMMES (UMR 5819), CIBEST, 17 rue des martyrs, F-38 000 Grenoble, France. d. Center for Natural Sciences and Humanities, Federal University of ABC – UFABC, 09210-580, Santo André, SP, Brazil e. BM30B/FAME beamline, ESRF, F-38043 Grenoble cedex 9, France f. Observatoire des Sciences de l’Univers de Grenoble, UMS 832 CNRS Université Grenoble Alpes, F-38041 Grenoble, France g. Institut de Chimie Moléculaire de Reims (ICMR, UMR CNRS 7312), Université de Reims Champagne-Ardenne, F- 51687 Reims Cedex 2, France

§ These two authors contributed equally to the work.

Supplementary Information

Content

Scheme S1. Synthesis of L

Figure S1. (+)ESI-MS spectrum of the Cu(II)L complex.

Figure S2. EPR signatures of Cu(II)L at different pH values.

Figure S3. (+)ESI-MS spectrum of the Cu(I)L complex.

Figure S4. EXAFS data for the Cu(I)L complex.

Figure S5. Cyclic voltamograms of O2 and ascorbate

Figure S6. EPR competition between Cu(II)Aβ1-16 and L. Linear combinations.

Figure S7. XANES competition beween Cu(I)Aβ1-16 and L. Linear combinations.

Figure S8. Kinetics of ascorbate consumption with peptide Aβ1-16.

Figure S9. Fluorescence kinetics of CCA experiments

1 Figure S10. 400 MHz H NMR spectrum in D2O at 300 K.

13 1 Figure S11. 100 MHz C NMR spectrum decoupled from H in D2O at 300 K.

~ 108 ~

Chapter II: Cu ion chelation

HN N O

H2N NHS O N HN O NH NH2 H α NHS N N + N O O β 5 O O H2N O O 4 NH NH2 NHS 2 HCl N NH N 2 HN O NH2

Scheme S1. Synthesis of L ≈ NTA(HisNH2)3 – reagents and conditions: DIEA, CH3CN/DMF, 11 %

Figure S1. (+)ESI-MS spectrum of L (100 µM) with equimolar Cu(II)SO4, in ammonium acetate buffer (20 mM, pH 6.9). * Sodium adducts. # Potassium adducts.

~ 109 ~

Chapter II: Cu ion chelation

Figure S2. EPR signatures of Cu(II)L at different pH values. [L] = 200 µM, [65Cu(II)] = 180 µM. 10% of glycerol was used as cryoprotectant. T = 110 K.

Figure S3. (+)ESI-MS spectrum of L (100 µM) with equimolar Cu(I)(CH3CN)4PF6, in ammonium acetate buffer (20 mM, pH 6.9).

~ 110 ~

Chapter II: Cu ion chelation

data fit

(k) χ

k*

4 6 8 10 12

k(Å-1)

FT fit

Im(FT) fit FT data Im(FT) data

1 2 3 4 R(Å)

Figure S4. EXAFS data for the Cu(I)L complex. Top. Experimental data (solid black curve) and simulated fit (dotted red line) of the filtered first shell of the EXAFS function and Down. the corresponding radial structure functions (RSF, not corrected phase shift). FT and Im(FT) are the magnitude and imaginary part of Fourier transforms, respectively.

~ 111 ~

Chapter II: Cu ion chelation

Figure S5. Cyclic voltamograms of O2 (Panel A) and ascorbate (Panel B). [O2] = ~ 0.2 mM, [Asc] = 1 mM in [phosphate buffer] = 100 mM at pH 7.1. Scan rate = 100 mV.s-1 ; WE = Glassy carbon, Ref = SCE, CE = Pt wire.

Figure S6. Panel A. EPR experiments of Cu(II)Aβ1-16 + Cu(II)L (black curve), linear combination: 0% Cu(II)Aβ1-16 + 100% Cu(II)L (blue curve), linear combination: 20% Cu(II)Aβ1-16 + 80% Cu(II)L (pink curve). The inset corresponds to a zoom of the first band. Panel B. EPR experiments of Cu(II)Aβ1-16 + Cu(II)L (black curve), linear combination: 5% Cu(II)Aβ1-16 + 95% Cu(II)L (blue curve), linear combination: 10% Cu(II)Aβ1-16 + 90% Cu(II)L (pink curve). The inset corresponds to a zoom of the first band. [L] = [Aβ1- 65 16] = [ Cu(II)] = 200 µM, [HEPES] = 50 mM, pH 7.1. 10% of glycerol was used as cryoprotectant. T = 110 K.

Only insignificant difference can be observed in the reproduction of the experimental competition spectrum using 0 -100 % ; 5 – 95 % and 10 % - 90 % as Cu(II)Aβ1-16– CuLH-1 ratio. We thus propose that in the competition spectrum there is 5 % +/- 5% of Cu(II)Aβ1-16.

~ 112 ~

Chapter II: Cu ion chelation

Figure S7. Panel A. XANES spectra of Cu(I)Aβ1-16 + Cu(I)L (black line), linear combination: 0% Cu(I)Aβ1-16 + 100% Cu(I)L (blue line), linear combination: 100% Cu(I)Aβ1-16 + 0% Cu(I)L (red line). Panel B. XANES spectra of Cu(I)Aβ1-16 + Cu(I)L (black line), linear combination: 0% Cu(I)Aβ1-16 + 100% Cu(I)L (blue line), linear combination: 50% Cu(I)Aβ1-16 + 50% Cu(I)L (red line). Panel C. XANES spectra of Cu(I)Aβ1-16 + Cu(I)L (black line), linear combination: 10% Cu(I)Aβ1-16 + 90% Cu(I)L (blue line), linear combination: 30% Cu(I)Aβ1-16 + 70% Cu(I)L (red line). Panel D. XANES spectra of Cu(I)Aβ1-16 + Cu(I)L (black line), linear combination: 15% Cu(I)Aβ1-16 + 85% Cu(I)L (blue line), linear combination: 20% Cu(I)Aβ1-16 + 80% Cu(I)L (red line), linear combination: 25% Cu(I)Aβ1-16 + 75% Cu(I)L (green line). The insets correspond to a zoom of the absorption. [L] = [Aβ1-16] = 1.00 mM, [Cu(II)] = 0.95 mM, [dithionite] = 10 mM, [HEPES] = 100 mM, pH 7.1. 10% of glycerol was used as cryoprotectant.

~ 113 ~

Chapter II: Cu ion chelation

Figure S8. Top: Kinetics of ascorbate consumption, followed by UV-visible spectroscopy at 265 with subtraction of the background signal at 800 nm. Panel A. Aβ1-16 + Cu(II) + Asc (black curve), L + Cu(II) + Asc (blue curve), Aβ1-16 + Cu(II) + L + Asc (red curve). Panel B. Cu(II) + Asc + Aβ1-16 + air (black curve), Cu(II) + Asc + L + air (blue curve), Cu(II) + Asc + Aβ1-16 + L + air (red curve). Panel C. Asc + Cu(II) + L (blue curve), Asc + Aβ1-16 + Cu(II) + L (red curve). [L] = [Aβ1-16] = 12 µM, [Cu(II)] = 10 µM, [Asc] = 100 µM, [HEPES] = 100 mM, pH 7.1. For the experiments from Panel B, all the solutions were deoxygenated by bubbling Argon and were added under a little overpressure of Argon in order to keep Cu under its +I oxidation state. Bottom: Panel A. UV-vis spectra of L (blue curve), L + Cu(II) (red curve), L + Cu(I) (green curve). [L] = 100 µM, [Cu(II)] = [Cu(I)] = 100 µM, [HEPES] = 100 mM, pH 7.1. To ease direct comparison, the absorbance intensity has been divided by a factor of ten since data were recorded at a ten-fold higher concentration to improve the signal over noise ratio. Panel B. Spectra corresponding to the experiment shown in Figure 7. Asc + Aβ40 (black curve), Asc + Aβ40 + Cu(II) at 720 s, just before the addition of L (dark grey curve), Asc + Aβ40 + Cu(II) + L at the end of the kinetic (light grey curve). [L] = [Aβ1-16] = 12 µM, [Cu(II)] = 10 µM, [Asc] = 100 µM, [HEPES] = 100 mM, pH 7.1.

~ 114 ~

Chapter II: Cu ion chelation

Figure S9. Fluorescence kinetics of CCA experiments of (a) Cu(II) + Asc, (b) Aβ1-16 + Cu(II) + Asc, (c) L + Cu(II) + Asc, (d) Aβ1-16 + Cu(II) + L + Asc, (e) Aβ1-16 + Cu(II) + Asc + L at 20 min. Asc was added 5 min after the beginning of the measurement. [L] = [Aβ1- 16] = 12 µM, [Cu(II)] = 10 µM, [CCA] = 500 µM, [Asc] = 500 µM, [phosphate buffer] = 50 mM, pH 7.1.

1 Figure S10. 400 MHz H NMR spectrum in D2O at 300 K.

~ 115 ~

Chapter II: Cu ion chelation

13 1 Figure S11. 100 MHz C NMR spectrum decoupled from H in D2O at 300 K.

~ 116 ~

Chapter II: Cu ion chelation

II-D Conclusion

In this part regarding the Cu chelation in the AD context, two proofs of concept have been detailed. The importance of the kinetic in the Cu removal from the Aβ peptide has been illustrated with two series of macrocyclic ligands. Indeed, if the redox process switching from Cu(II)-Aβ to Cu(I)-Aβ is faster than the Cu removal by the ligand, Cu ions stay bound to Aβ, leading to the production of ROS and the stabilisation of species such as the toxic oligomers. Then, the importance of targeting Cu(I) and Cu(II) ions in the chelatotherapy has been shown using the ligand L. As the redox state of Cu ions in the synaptic cleft is still unknown, it could be more useful to target both redox states. A caution has to be paid on the redox silencing of the ROS production by the Cu complex formed. A more secure approach in the chelatotherapy could be the use of a ligand with a fast kinetic in the removal of Cu ions from Aβ and able to chelate both Cu(I) and Cu(II). As it could chelate Cu ions faster than the redox process and as it could chelate both Cu(I) and Cu(II), it should be able to remove Cu ions from Aβ in vivo.

Nevertheless, one can wonder if a ligand, with a fast Cu chelation kinetic, in vivo, could not remove Cu ions after its administration and before reaching the synaptic cleft. This could have severe negative effects. In vivo, another issue for a Cu(I)/Cu(II) ligand can be the removal of Cu(I) from important metalloproteins found inside the cells, as the ligand is able to remove Cu(I). It is important that the ligand does not exhibit a too high affinity for Cu(I) (as for the Cu(II) affinity constant), compared to the affinity constant of the intra-cellular metalloproteins such as the metallothioneins or the Cu- chaperons. Furthermore, adding a targeting Aβ moiety to such a ligand could help in the removal of the “AD Cu ions”.

Another important point could be the metallophoric capabilities of a Cu(I)/Cu(II) ligand. Indeed, as previously explained, there is a dyshomeostasis of Cu ions the brain. The final idea of the chelatotherapy is not only the removal of Cu ions from Aβ but also their intra-cellular redistribution. In the literature, the metallophores are Cu(II) ligands which, under a reducing medium as the intra- cellular medium, get reduced and release the Cu(I). However, a Cu(I) and Cu(II) ligand will form quite stable Cu(II) but also Cu(I) complexes. It can be more difficult to release the Cu(I) from such a stable complex. The design of such a ligand has to take into account that a Cu-protein in the intra-cellular medium has to be able to remove Cu(I) from the Cu(I)-L formed, in order to re-equilibrate the Cu homeostasis.

~ 117 ~

Chapter II: Cu ion chelation

References

1. Rowinska-Zyrek, M.; Salerno, M.; Kozlowski, H., Neurodegenerative diseases – Understanding their molecular bases and progress in the development of potential treatments. Coord. Chem. Rev. 2015, 284, 298–312. 2. Xia, N.; Liu, L., Metallothioneins and Synthetic Metal Chelators as Potential Therapeutic Agents for Removal of Aberrant Metal Ions from Metal-AB Species. Mini Rev. Med. Chem. 2014, 14 (3), 271- 281. 3. Robert, A.; Liu, Y.; Nguyen, M.; Meunier, B., Regulation of copper and iron homeostasis by metal chelators: a possible chemotherapy for Alzheimer's disease. Acc. Chem. Res. 2015, 48 (5), 1332- 1339. 4. Santos, M. A.; Chand, K.; Chaves, S., Recent progress in multifunctional metal chelators as potential drugs for Alzheimer’s disease. Coord. Chem. Rev. 2016, 327 –328, 287–303. 5. Scott, L. E.; Orvig, C., Medicinal inorganic chemistry approaches to passivation and removal of aberrant metal ions in disease. Chem. Rev. 2009, 109 (10), 4885-4910. 6. Hung, L. W.; Barnham, K. J., Modulating metals as a therapeutic strategy for Alzheimer's disease. 2012, 4 (8), 955-969. 7. Gnjec, A.; Fonte, J. A.; Atwood, C.; Martins, R. N., Transition metal chelator therapy – A potential treatment for alzheimer’s disease? Front. Biosci. 2002, 7, 1016-1023. 8. Budimir, A., Metal ions, Alzheimer’s disease and chelation therapy. Acta Pharm. 2011, 61, 1- 14. 9. Rodríguez-Rodríguez, C.; Telpoukhovskaia, M.; Orvig, C., The art of building multifunctional metal-binding agents from basic molecular scaffolds for the potential application in neurodegenerative diseases. Coord. Chem. Rev. 2012, 256 (19), 2308-2332. 10. Cuajungco, M. P.; Fagét, K. Y.; Huang, X.; Tanzi, R. E.; Bush, A. I., Metal chelation as a potential therapy for Alzheimer's disease. Ann. N.Y. Acad. Sci. 2000, 920, 292-304. 11. Bolognin, S.; Drago, D.; Messori, L.; Zatta, P., Chelation therapy for neurodegenerative diseases. Med Res Rev. 2009, 29 (4), 547-570. 12. Wezynfeld, N. E.; Stefaniak, E.; Stachucy, K.; Drozd, A.; Płonka, D.; Drew, S. C.; Krężel, A.; Bal, W., Resistance of Cu(Aβ4-16) to Copper Capture by Metallothionein-3 Supports a Function for the Aβ4- 42 Peptide as a Synaptic Cu(II) Scavenger. Angew. Chem. Int. Ed. 2016, 55 (29), 8235-8238. 13. Mital, M.; Wezynfeld, N. E.; Fraczyk, T.; Wiloch, M. Z.; Wawrzyniak, U. E.; Bonna, A.; Tumpach, C.; Barnham, K. J.; Haigh, C. L.; Bal, W.; Drew, S. C., A Functional Role for Aβ in Metal Homeostasis? N- Truncation and High-Affinity Copper Binding. Angew. Chem. Int. Ed. 2015, 54 (36), 10460-10464. 14. Gabathuler, R., Approaches to transport therapeutic drugs across the blood–brain barrier to treat brain diseases. Neurobiol. Dis. 2010, 37 (1), 48-57. 15. Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J., Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Adv. Drug Deliv. Rev. 2001, 46 (1), 3-26. 16. Savelieff, M. G.; DeToma, A. S.; Derrick, J. S.; Lim, M. H., The Ongoing Search for Small Molecules to Study Metal-Associated Amyloid-β Species in Alzheimer’s Disease. Acc. Chem. Res. 2014, 47 (8), 2475–2482. 17. Bush, A. I.; Tanzi, R. E., Therapeutics for Alzheimer’s Disease Based on the Metal Hypothesis. Neurotherapeutics 2008, 5, 421–432. 18. Gaeta, A.; Hider, R. C., The crucial role of metal ions in neurodegeneration: the basis for a promising therapeutic strategy. Br. J. Pharmacol. 2005, 146 (8), 1041-1059. 19. Derrick, J. S.; Lim, M. H., Tools of the Trade: Investigations into Design Strategies of Small Molecules to Target Components in Alzheimer's Disease. ChemBioChem 2015, 16 (6), 887-898. 20. Masui, H., Metalloaromaticity. Coord. Chem. Rev. 2001, 219, 957-992.

~ 118 ~

Chapter II: Cu ion chelation

21. Rimola, A.; Alí-Torres, J.; Rodríguez-Rodríguez, C.; Poater, J.; Matito, E.; Solà, M.; Sodupe, M., Ab Initio Design of Chelating Ligands Relevant to Alzheimer’s Disease: Influence of Metalloaromaticity. J. Phys. Chem. A. 2011, 115 (45), 12659-12666. 22. Perez, L. R.; Franz, K. J., Minding metals: Tailoring multifunctional chelating agents for neurodegenerative disease. Dalton Trans. 2010, 39 (9), 2177-2187. 23. Beck, M. W.; Pithadia, A. S.; DeToma, A. S.; Korshavn, K. J.; Lim, M. H., Ligand Design to Target and Modulate Metal–Protein Interactions in Neurodegenerative Diseases. In Ligand Design in Medicinal Inorganic Chemistry, Storr, T., Ed. Wiley: 2014; pp 256-286. 24. Sahni, J. K.; Doggui, S.; Ali, J.; Baboota, S.; Dao, L.; Ramassamy, C., Neurotherapeutic applications of nanoparticles in Alzheimer's disease. J. Controlled Release 2011, 152 (2), 208-231. 25. Liu, G.; Men, P.; Kudo, W.; Perry, G.; Smith, M. A., Nanoparticle–chelator conjugates as inhibitors of amyloid-β aggregation and neurotoxicity: A novel therapeutic approach for Alzheimer disease. Neurosci. Lett. 2009, 455 (3), 187-190. 26. Cui, Z.; Lockman, P. R.; Atwood, C. S.; Hsu, C. H.; Gupte, A.; Allen, D. D.; Mumper, R. J., Novel D-penicillamine carrying nanoparticles for metal chelation therapy in Alzheimer's and other CNS diseases. Eur. J. Pharm. Biopharm. 2005, 59 (2), 263-272. 27. Berk, C.; Paul, G.; Sabbagh, M., Investigational drugs in Alzheimer's disease: current progress. Expert Opin. Investig. Drugs 2014, 23 (6), 837-846. 28. Guzior, N.; Więckowska, A.; Panek, D.; Malawska, B., Recent Development of Multifunctional Agents as Potential Drug Candidates for the Treatment of Alzheimer's Disease. Curr. Med. Chem. 2015, 22 (3), 373-404. 29. Braymer, J. J.; DeToma, A. S.; Choi, J.-S.; Ko, K. S.; Lim, M. H., Recent Development of Bifunctional Small Molecules to Study Metal-Amyloid-β Species in Alzheimer’s Disease. Int. J. Alzheimers Dis. 2011, 2011, 1-9. 30. Dickens, M. G.; Franz, K. J., A Prochelator Activated by Hydrogen Peroxide Prevents Metal- Induced Amyloid β Aggregation. ChemBioChem 2010, 11 (1), 59-62. 31. Folk, D. S.; Franz, K. J., A Prochelator Activated by β−Secretase Inhibits Aβ Aggregation and Suppresses Copper-Induced Reactive Oxygen Species Formation. J. Am. Chem. Soc. 2010, 132, 4994– 4995. 32. Hung, Y. H.; Bush, A. I.; Cherny, R. A., Copper in the brain and Alzheimer’s disease. J. Biol. Inorg. Chem. 2010, 15, 61–76. 33. Squitti, R.; Salustri, C., Agents complexing copper as a therapeutic strategy for the treatment of Alzheimer's disease. Curr. Alzheimer Res. 2009, 6 (6), 476-487. 34. Barnham, K. J.; Bush, A. I., Biological metals and metal-targeting compounds in major neurodegenerative diseases. Chem. Soc. Rev. 2014, 43 (19), 6727-6749. 35. Ayton, S.; Lei, P.; Bush, A. I., Biometals and their therapeutic implications in Alzheimer's disease. Neurotherapeutics 2015, 12 (1), 109-120. 36. Bush, A. I., Drug Development Based on the Metals Hypothesis of Alzheimer’s Disease. J. Alzheimers Dis. 2008, 15, 223-240. 37. Crouch, P. J.; Barnham, K. J., Therapeutic Redistribution of Metal Ions To Treat Alzheimer's Disease. Acc. Chem. Res. 2012, 45 (9), 1604–1611. 38. Bareggi, S. R.; Cornelli, U., Clioquinol: Review of its Mechanisms of Action and Clinical Uses in Neurodegenerative Disorders. CNS Neurosci Ther 2012, 18, 41-46. 39. Bush, A. I., Metal complexing agents as therapies for Alzheimer’s disease. Neurobiol. Aging 2002, 23, 1031-1038. 40. Zatta, P.; Drago, D.; Bolognin, S.; Sensi, S. L., Alzheimer’s disease, metal ions and metal homeostatic therapy. Trends Pharmacol. Sci. 2009, 30 (7), 346-355. 41. Cahoon, L., The curious case of clioquinol. Nat. Med. 2009, 15 (4), 356-359. 42. Bolognin, S.; Zatta, P.; Drago, D.; Tognon, G.; Parnigotto, P. P.; Ricchelli, F., Mutual stimulation of beta-amyloid fibrillogenesis by clioquinol and divalent metals. Neuromol. Med. 2008, 10 (4), 322- 332.

~ 119 ~

Chapter II: Cu ion chelation

43. Di Vaira, M.; Bazzicalupi, C.; Orioli, P.; Messori, L.; Bruni, B.; Zatta, P., Clioquinol, a drug for Alzheimer's disease specifically interfering with brain metal metabolism: structural characterization of its zinc(II) and copper(II) complexes. Inorg. Chem. 2004, 43 (13), 3795-3797. 44. Budimir, A.; Humbert, N.; Elhabiri, M.; Osinska, I.; Birus, M.; Albrecht-Gary, A. M., Hydroxyquinoline based binders: promising ligands for chelatotherapy? J. Inorg. Biochem. 2011, 105 (3), 490-496. 45. Pushie, M. J.; Nienaber, K. H.; Summers, K. L.; Cotelesage, J. J. H.; Ponomarenko, O.; Nichol, H. K.; Pickering, I. J.; George, G. N., The solution structure of the copper clioquinol complex. J. Inorg. Biochem. 2014, 133, 50-56. 46. Raman, B.; Ban, T.; Yamaguchi, K.-i.; Sakai, M.; Kawai, T.; Naiki, H.; Goto, Y., Metal Ion- dependent Effects of Clioquinol on the Fibril Growth of an Amyloid β Peptide. J. Biol. Chem. 2005, 280 (16), 16157-16162. 47. Mancino, A. M.; Hindo, S. S.; Kochi, A.; Lim, M. H., Effects of Clioquinol on Metal-Triggered Amyloid-β Aggregation Revisited. Inorg. Chem. 2009, 48 (20), 9596-9598. 48. Padmanabhan, G.; Becue, I.; Smith, J., Analytical profiles of drug substances. Academic Press, New York: 1989; pp 57-90. 49. Cherny, R. A.; Atwood, C. S.; Xilinas, M. E.; Gray, D. N.; Jones, W. D.; McLean, C. A.; Barnham, K. J.; Volitakis, I.; Fraser, F. W.; Kim, Y.-S.; Huang, X.; Goldstein, L. E.; Moir, R. D.; Lim, J. T.; Beyreuther, K.; Zheng, H.; Tanzi, R. E.; Masters, C. L.; Bush, A. I., Treatment with a copper-zinc chelator markedly and rapidly inhibits beta-amyloid accumulation in Alzheimer's disease transgenic mice. Neuron 2001, 30 (3), 665-676. 50. Grossi, C.; Francese, S.; Casini, A.; Rosi, M. C.; Luccarini, I.; Fiorentini, A.; Gabbiani, C.; Messori, L.; Moneti, G.; Casamenti, F., Clioquinol decreases amyloid-beta burden and reduces working memory impairment in a transgenic mouse model of Alzheimer's disease. J. Alzheimers Dis. 2009, 17 (2), 423- 440. 51. Schafer, S.; Pajonk, F. G.; Multhaup, G.; Bayer, T. A., Copper and clioquinol treatment in young APP transgenic and wild-type mice: effects on life expectancy, body weight, and metal-ion levels. Journal of molecular medicine 2007, 85 (4), 405-413. 52. Regland, B.; Lehmann, W.; Abedini, I.; Blennow, K.; Jonsson, M.; Karlsson, I.; Sjogren, M.; Wallin, A.; Xilinas, M.; Gottfries, C. G., Treatment of Alzheimer's disease with clioquinol. Dement. Geriatr. Cogn. Dis. 2001, 12 (6), 408-14. 53. Ibach, B.; Haen, E.; Marienhagen, J.; Hajak, G., Clioquinol treatment in familiar early onset of Alzheimer's disease: a case report. Pharmacopsychiatry 2005, 38 (4), 178-179. 54. Ritchie, C. W.; Bush, A. I.; Mackinnon, A.; Macfarlane, S.; Mastwyk, M.; MacGregor, L.; Kiers, L.; Cherny, R. A.; Li, Q.-X.; Tammer, A.; Carrington, D.; Mavros, C.; Volitakis, I.; Xilinas, M. E.; Ames, D.; Davis, S.; Beyreuther, K.; Tanzi, R. E.; Masters, C. L., Metal-protein attenuation with iodochlorhydroxyquin (clioquinol) targeting Aβ amyloid deposition and toxicity in alzheimer disease: A pilot phase 2 clinical trial. Arch. Neurol. 2003, 60 (12), 1685-1691. 55. Ritchie, C. W.; Bush, A. I.; Masters, C. L., Metal-protein attenuating compounds and Alzheimer's disease. Expert Opin. Investig. Drugs 2004, 13 (12), 1585-1592. 56. Rosenberg, R. N., Translational research on the way to effective therapy for Alzheimer disease. Arch. Gen. Psychiatry 2005, 62 (11), 1186-1192. 57. Kenche, V. B.; Barnham, K. J., Alzheimer’s disease & metals: therapeutic opportunities. Br. J. Pharmacol. 2011, 163, 211-219. 58. Treiber, C.; Simons, A.; Strauss, M.; Hafner, M.; Cappai, R.; Bayer, T. A.; Multhaup, G., Clioquinol Mediates Copper Uptake and Counteracts Copper Efflux Activities of the Amyloid Precursor Protein of Alzheimer's Disease. J. Biol. Chem. 2004, 279 (50), 51958–51964. 59. Mital, M.; Zawisza, I. A.; Wiloch, M. Z.; Wawrzyniak, U. E.; Kenche, V.; Wróblewski, W.; Bal, W.; Drew, S. C., Copper Exchange and Redox Activity of a Prototypical 8-Hydroxyquinoline: Implications for Therapeutic Chelation. Inorg. Chem. 2016, 55 (15), 7317-7319. 60. Barnham, K. J.; Gautier, E. C. L.; Kok, G. B.; Krippner, G., 8-hydroxy quinoline derivatives. In WO 2004007461 A1, Google Patents: 2004.

~ 120 ~

Chapter II: Cu ion chelation

61. Fu, C. L.; Hsu, L. S.; Liao, Y. F.; Hu, M. K., New Hydroxyquinoline-Based Derivatives as Potent Modulators of Amyloid-beta Aggregations. Arch. Pharm. 2016, 349 (5), 327-341. 62. Adlard, P. A.; Cherny, R. A.; Finkelstein, D. I.; Gautier, E.; Robb, E.; Cortes, M.; Volitakis, I.; Liu, X.; Smith, J. P.; Perez, K.; Laughton, K.; Li, Q.-X.; Charman, S. A.; Nicolazzo, J. A.; Wilkins, S.; Deleva, K.; Lynch, T.; Kok, G.; Ritchie, C. W.; Tanzi, R. E.; Cappai, R.; Masters, C. L.; Barnham, K. J.; Bush, A. I., Rapid restoration of cognition in Alzheimer's transgenic mice with 8-hydroxy quinoline analogs is associated with decreased interstitial Abeta. Neuron 2008, 59 (1), 43-55. 63. Lannfelt, L.; Blennow, K.; Zetterberg, H.; Batsman, S.; Ames, D.; Harrison, J.; Masters, C. L.; Targum, S.; Bush, A. I.; Murdoch, R.; Wilson, J.; Ritchie, C. W., Safety, efficacy, and biomarker findings of PBT2 in targeting Aβ as a modifying therapy for Alzheimer's disease: a phase IIa, double-blind, randomised, placebo-controlled trial. Lancet Neurol. 2008, 7 (9), 779-786. 64. Faux, N. G.; Ritchie, C. W.; Gunn, A.; Rembach, A.; Tsatsanis, A.; Bedo, J.; Harrison, J.; Lannfelt, L.; Blennow, K.; Zetterberg, H.; Ingelsson, M.; Masters, C. L.; Tanzi, R. E.; Cummings, J. L.; Herd, C. M.; Bush, A. I., PBT2 rapidly improves cognition in Alzheimer's Disease: additional phase II analyses. J. Alzheimers Dis. 2010, 20 (2), 509-516. 65. Relkin, N. R., Testing the mettle of PBT2 for Alzheimer's disease. The Lancet. Neurology 2008, 7 (9), 762-763. 66. Oliveri, V.; Vecchio, G., 8-Hydroxyquinolines in medicinal chemistry: A structural perspective. Eur. J. Med. Chem. 2016, 120, 252-274. 67. Hindo, S. S.; Mancino, A. M.; Braymer, J. J.; Liu, Y.; Vivekanandan, S.; Ramamoorthy, A.; Lim, M. H., Small Molecule Modulators of Copper-Induced Aβ Aggregation. J. Am. Chem. Soc. 2009, 131 (46), 16663-16665. 68. Martínez, A.; Alcendor, R.; Rahman, T.; Podgorny, M.; Sanogo, I.; McCurdy, R., Ionophoric polyphenols selectively bind Cu2+, display potent antioxidant and anti-amyloidogenic properties, and are non-toxic toward Tetrahymena thermophila. Biorg. Med. Chem. 2016, 24 (16), 3657-3670. 69. Barcia, E.; Salama, A.; Fernández-Carballido, A.; Negro, S., Protective effects of clioquinol on human neuronal-like cells: a new formulation of clioquinol-loaded PLGA microspheres for Alzheimer’s disease. Journal of Drug Targeting 2011, 19 (8), 637-646. 70. Gomes, L. M.; Vieira, R. P.; Jones, M. R.; Wang, M. C.; Dyrager, C.; Souza-Fagundes, E. M.; Da Silva, J. G.; Storr, T.; Beraldo, H., 8-Hydroxyquinoline Schiff-base compounds as antioxidants and modulators of copper-mediated Aβ peptide aggregation. J. Inorg. Biochem. 2014, 139, 106-116. 71. Crouch, P. J.; Savva, M. S.; Hung, L. W.; Donnelly, P. S.; Mot, A. I.; Parker, S. J.; Greenough, M. A.; Volitakis, I.; Adlard, P. A.; Cherny, R. A.; Masters, C. L.; Bush, A. I.; Barnham, K. J.; White, A. R., The Alzheimer's therapeutic PBT2 promotes amyloid-β degradation and GSK3 phosphorylation via a metal chaperone activity. J. Neurochem. 2011, 119, 220–230. 72. Nguyen, M.; Robert, A.; Soumia-Saquet, A.; Vendier, L.; Meunier, B., Characterization of new specific copper chelators as potential drugs for the treatment of Alzheimer's disease. Chem. Eur. J. 2014, 20 (22), 6771-6785. 73. Nguyen, M.; Rechignat, L.; Robert, A.; Meunier, B., The Necessity of Having a Tetradentate Ligand to Extract Copper(II) Ions from Amyloids. ChemistryOpen 2015, 4 (1), 27-31. 74. Nguyen, M.; Bijani, C.; Martins, N.; Meunier, B.; Robert, A., Transfer of Copper from an Amyloid to a Natural Copper-Carrier Peptide with a Specific Mediating Ligand. Chem. Eur. J. 2015, 21 (47), 17085-17090. 75. Ceccom, J.; Coslédan, F.; Halley, H.; Francès, B.; Lassalle, J. M.; Meunier, B., Copper Chelator Induced Efficient Episodic Memory Recovery in a Non-Transgenic Alzheimer’s Mouse Model. PLoS One 2012, 7 (8), e43105. 76. Deraeve, C.; Boldron, C.; Maraval, A.; Mazarguil, H.; Gornitzka, H.; Vendier, L.; Pitié, M.; Meunier, B., Preparation and Study of New Poly-8-Hydroxyquinoline Chelators for an anti-Alzheimer Strategy. Chem. Eur. J. 2008, 14 (2), 682-696. 77. Kenche, V. B.; Zawisza, I. A.; Masters, C. L.; Bal, W.; Barnham, K. J.; Drew, S. C., Mixed ligand Cu2+ complexes of a model therapeutic with Alzheimer's amyloid-β peptide and monoamine neurotransmitters. Inorg. Chem. 2013, 52 (8), 4303-4318.

~ 121 ~

Chapter II: Cu ion chelation

78. Choi, J.-S.; Braymer, J. J.; Nanga, R. P. R.; Ramamoorthy, A.; Lim, M. H., Design of small molecules that target metal-Aβ species and regulate metal-induced Aβ aggregation and neurotoxicity. PNAS 2010, 107 (51), 21990-21995. 79. Beck, M. W.; Oh, S. B.; Kerr, R. A.; Lee, H. J.; Kim, S. H.; Kim, S.; Jang, M.; Ruotolo, B. T.; Lee, J.- Y.; Lim, M. H., A rationally designed small molecule for identifying an in vivo link between metal– amyloid-β complexes and the pathogenesis of Alzheimer's disease. Chemical Science 2015, 6 (3), 1879- 1886. 80. Beck, M. W.; Derrick, J. S.; Kerr, R. A.; Oh, S. B.; Cho, W. J.; Lee, S. J. C.; Ji, Y.; Han, J.; Tehrani, Z. A.; Suh, N.; Kim, S.; Larsen, S. D.; Kim, K. S.; Lee, J.-Y.; Ruotolo, B. T.; Lim, M. H., Structure-mechanism- based engineering of chemical regulators targeting distinct pathological factors in Alzheimer’s disease. Nat. Commun. 2016, 7, 13115-13127. 81. Lee, S.; Zheng, X.; Krishnamoorthy, J.; Savelieff, M. G.; Park, H. M.; Brender, J. R.; Kim, J. H.; Derrick, J. S.; Kochi, A.; Lee, H. J.; Kim, C.; Ramamoorthy, A.; Bowers, M. T.; Lim, M. H., Rational Design of a Structural Framework with Potential Use to Develop Chemical Reagents That Target and Modulate Multiple Facets of Alzheimer’s Disease. J. Am. Chem. Soc. 2014, 136 (1), 299-310. 82. Geng, J.; Li, M.; Wu, L.; Ren, J.; Qu, X., Liberation of Copper from Amyloid Plaques: Making a Risk Factor Useful for Alzheimer’s Disease Treatment. J. Med. Chem. 2012, 55, 9146−9155. 83. Rodríguez-Rodríguez, C.; Sánchez de Groot, N.; Rimola, A.; Álvarez-Larena, Á.; Lloveras, V.; Vidal-Gancedo, J.; Ventura, S.; Vendrell, J.; Sodupe, M.; González-Duarte, P., Design, Selection, and Characterization of Thioflavin-Based Intercalation Compounds with Metal Chelating Properties for Application in Alzheimer’s Disease. J. Am. Chem. Soc. 2009, 131 (4), 1436-1451. 84. Harford, C.; Sarkar, B., Amino Terminal Cu(II)- and Ni(II)-Binding (ATCUN) Motif of Proteins and Peptides: Metal Binding, DNA Cleavage, and Other Properties. Acc. Chem. Res. 1997, 30 (3), 123-130. 85. Lakatos, A.; Gyurcsik, B.; Nagy, N. V.; Csendes, Z.; Weber, E.; Fulop, L.; Kiss, T., Histidine-rich branched peptides as Cu(ii) and Zn(ii) chelators with potential therapeutic application in Alzheimer's disease. Dalton Trans. 2012, 41 (6), 1713-1726. 86. Jensen, M.; Canning, A.; Chiha, S.; Bouquerel, P.; Pedersen, J. T.; Østergaard, J.; Cuvillier, O.; Sasaki, I.; Hureau, C.; Faller, P., Inhibition of Cu-Amyloid-β by using Bifunctional Peptides with β-Sheet Breaker and Chelator Moieties. Chem. Eur. J. 2012, 18 (16), 4836-4839. 87. Conte-Daban, A.; Boff, B.; Candido, A.; Montes Aparicio, C.; Gateau, C.; Lebrun, C.; Cerchiaro, G.; Kieffer, I.; Sayen, S.; Guillon, E.; Delangle, P.; Hureau, C., A trishistidine pseudopeptide with ability to remove both Cu(I) and Cu(II) from the amyloid-beta peptide and to stop the associated ROS formation. Chem. Eur. J. 2017, In press. 88. Caballero, A. B.; Terol-Ordaz, L.; Espargaró, A.; Vázquez, G.; Nicolás, E.; Sabaté, R.; Gamez, P., Histidine-Rich Oligopeptides To Lessen Copper-Mediated Amyloid-β Toxicity. Chem. Eur. J. 2016, 22 (21), 7268-7280. 89. Hu, X.; Zhang, Q.; Wang, W.; Yuan, Z.; Zhu, X.; Chen, B.; Chen, X., Tripeptide GGH as the Inhibitor of Copper-Amyloid-β-Mediated Redox Reaction and Toxicity. ACS Chem. Neurosci. 2016, 7 (9), 1255-1263. 90. Quintanova, C.; Keri, R. S.; Chaves, S.; Santos, M. A., Copper(II) complexation of tacrine hybrids with potential anti-neurodegenerative roles. J. Inorg. Biochem. 2015, 151, 58-66. 91. Chen, T.; Wang, X.; He, Y.; Zhang, C.; Wu, Z.; Liao, K.; Wang, J.; Guo, Z., Effects of Cyclen and Cyclam on Zinc(II)- and Copper(II)-Induced Amyloid β-Peptide Aggregation and Neurotoxicity. Inorg. Chem. 2009, 48 (13), 5801-5809. 92. Wu, W.-h.; Lei, P.; Liu, Q.; Hu, J.; Gunn, A. P.; Chen, M.-s.; Rui, Y.-f.; Su, X.-y.; Xie, Z.-p.; Zhao, Y.-F.; Bush, A. I.; Li, Y.-m., Sequestration of Copper from β-Amyloid Promotes Selective Lysis by Cyclen- Hybrid Cleavage Agents. J. Biol. Chem. 2008, 283 (46), 31657-31664. 93. Yang, Y.; Chen, T.; Zhu, S.; Gu, X.; Jia, X.; Lu, Y.; Zhu, L., Two macrocyclic polyamines as modulators of metal-mediated Aβ40 aggregation. Integr. Biol. 2015, 7 (6), 655-662. 94. Kaden, D.; Bush, A. I.; Danzeisen, R.; Bayer, T. A.; Multhaup, G., Disturbed Copper Bioavailability in Alzheimer’s Disease. Int. J. Alzheimers Dis. 2011, 2011.

~ 122 ~

Chapter II: Cu ion chelation

95. Storr, T.; Merkel, M.; Song-Zhao, G. X.; Scott, L. E.; Green, D. E.; Bowen, M. L.; Thompson, K. H.; Patrick, B. O.; Schugar, H. J.; Orvig, C., Synthesis, Characterization, and Metal Coordinating Ability of Multifunctional Carbohydrate-Containing Compounds for Alzheimer’s Therapy. J. Am. Chem. Soc. 2007, 129, 7453-7463. 96. Conte-Daban, A.; Day, A.; Faller, P.; Hureau, C., How Zn can impede Cu detoxification by chelating agents in Alzheimer's disease: a proof-of-concept study. Dalton Trans. 2016, 45 (39), 15671- 15678. 97. Noël, S.; Perez, F.; Pedersen, J. T.; Alies, B.; Ladeira, S.; Sayen, S.; Guillon, E.; Gras, E.; Hureau, C., A new water-soluble Cu(II) chelator that retrieves Cu from Cu(amyloid-β) species, stops associated ROS production and prevents Cu(II)-induced Aβ aggregation. J. Inorg. Biochem. 2012, 117, 322-325. 98. Lakatos, A.; Zsigo, E.; Hollender, D.; Nagy, N. V.; Fulop, L.; Simon, D.; Bozso, Z.; Kiss, T., Two pyridine derivatives as potential Cu(ii) and Zn(ii) chelators in therapy for Alzheimer's disease. Dalton Trans. 2010, 39 (5), 1302-1315. 99. Atrian-Blasco, E.; Cerrada, E.; Conte-Daban, A.; Testemale, D.; Faller, P.; Laguna, M.; Hureau, C., Copper(i) targeting in the Alzheimer's disease context: a first example using the biocompatible PTA ligand. Metallomics 2015, 7 (8), 1229-1232. 100. Donnelly, P. S.; Caragounis, A.; Du, T.; Laughton, K.; Volitakis, I.; Cherny, R. A.; Sharples, R. A.; Hill, A. F.; Li, Q.-X.; Masters, C. L.; Barnham, K. J.; White, A. R., Selective intracellular release of copper and zinc ions from bis(thiosemicarbazonato) complexes reduces levels of Alzheimer disease amyloid- beta peptide. J. Biol. Chem. 2008, 283 (8), 4568-4577. 101. Crouch, P. J.; Hung, L. W.; Adlard, P. A.; Cortes, M.; Lal, V.; Filiz, G.; Perez, K. A.; Nurjono, M.; Caragounis, A.; Du, T.; Laughton, K.; Volitakis, I.; Bush, A. I.; Li, Q. X.; Masters, C. L.; Cappai, R.; Cherny, R. A.; Donnelly, P. S.; White, A. R.; Barnham, K. J., Increasing Cu bioavailability inhibits Aβ oligomers and tau phosphorylation. PNAS 2009, 106 (2), 381-386. 102. Sanabria-Castro, A.; Alvarado-Echeverría, I.; Monge-Bonilla, C., Molecular Pathogenesis of Alzheimer's Disease: An Update. Ann. Neurosci. 2017, 24 (1), 46-54. 103. Savelieff, M. G.; Lee, S.; Liu, Y.; Lim, M. H., Untangling Amyloid-β, Tau, and Metals in Alzheimer’s Disease. ACS Chem. Biol. 2014, 8, 856-865. 104. Atrian-Blasco, E.; Gonzalez, P.; Santoro, A.; Alies, B.; Faller, P.; Hureau, C., Cu and Zn coordination to amyloids: a chemistry of pathological importance ? Coord. Chem. Rev. 2017. 105. Ono, K., Alzheimer's disease as oligomeropathy. Neurochem. Int. 2017, 17, 30331-30335. 106. Solomon, E. I.; Heppner, D. E.; Johnston, E. M.; Ginsbach, J. W.; Cirera, J.; Qayyum, M.; Kieber- Emmons, M. T.; Kjaergaard, C. H.; Hadt, R. G.; Tian, L., Copper active sites in biology. Chem. Rev. 2014, 114 (7), 3659-3653. 107. Hureau, C., Coordination of redox active metal ions to the APP and to the amyloid-β peptides involved in AD. Part 1: an overview. Coord. Chem. Rev. 2012, 256 (19-20), 2164-2174. 108. Cheignon, C.; Tomas, M.; Bonnefont-Rousselot, D.; Faller, P.; Hureau, C.; Collin, F., Oxidative stress and the amyloid beta peptide in Alzheimer’s Disease. Redox Biology 2017, accepted. 109. Santos, M. A.; Chand, K.; Chaves, S., Recent progress in multifunctional metal chelators as potential drugs for Alzheimer’s disease. Coord. Chem. Rev. 2016, 327-328, 287-303. 110. Barnham, K. J.; Bush, A. I., Biological metals and metal-targeting compounds in major neurodegenerative diseases. Chem. Soc. Rev. 2014, 43, 6727-6749. 111. Atrian-Blasco, E.; Conte-Daban, A.; Hureau, C., Mutual interference of Cu and Zn ions in Alzheimer’s disease: perspectives at the molecular level. Dalton Transactions 2017, in press. 112. Chen, T.; Wang, X.; He, Y.; Zhang, C.; Wu, Z.; Liao, K.; Wang, J.; Zijian, G., Effects of Cyclen and Cyclam on Zinc(II)- and Copper(II)-Induced Amyloid β-Peptide Aggregation and Neurotoxicity. Inorg. Chem. 2009, 48 (13), 5801-5809. 113. Banerjee, S. R.; Pullambhatla, M.; Foss, C. A.; Nimmagadda, S.; Ferdani, R.; Anderson, C. J.; Mease, R. C.; Pomper, M. G., 64Cu-Labeled Inhibitors of Prostate-Specific Membrane Antigen for PET Imaging of Prostate Cancer. J. Med. Chem. 2014, 57 (6), 2657-2669.

~ 123 ~

Chapter II: Cu ion chelation

114. Stasiuk, G. J.; Long, N. J., The ubiquitous DOTA and its derivatives: the impact of 1,4,7,10- tetraazacyclododecane-1,4,7,10-tetraacetic acid on biomedical imaging. Chem. Commun. 2013, 49 (27), 2732-2746. 115. Wong, E. H.; Weisman, G. R.; Hill, D. C.; Reed, D. P.; Rogers, M. E.; Condon, J. S.; Fagan, M. A.; Calabrese, J. C.; Lam, K.-C.; Guzei, I. A.; Rheingold, A. L., Synthesis and Characterization of Cross-Bridged Cyclams and Pendant-Armed Derivatives and Structural Studies of Their Copper(II) Complexes. J. Am. Chem. Soc. 2000, 122 (43), 10561-10572. 116. Terreno, E.; Castelli, D. D.; Viale, A.; Aime, S., Challenges for Molecular Magnetic Resonance Imaging. Chem. Rev. 2010, 110 (5), 3019-3042. 117. Izatt, R. M.; Pawlak, K.; Bradshaw, J. S.; Bruening, R. L., Thermodynamic and kinetic data for macrocycle interaction with cations, anions, and neutral molecules. Chem. Rev. 1995, 95 (7), 2529- 2586. 118. Delgado, R.; Felix, V.; Lima, L. M. P.; Price, D. W., Metal complexes of cyclen and cyclam derivatives useful for medical applications: a discussion based on thermodynamic stability constants and structural data. Dalton Trans. 2007, (26), 2734-2745. 119. Burai, L.; Fabian, I.; Kiraly, R.; Szilagyi, E.; Brucher, E., Equilibrium and kinetic studies on the formation of the lanthanide(III) complexes, [Ce(dota)]- and [Yb(dota)]- (H4dota[space]=[space]1,4,7,10- tetraazacyclododecane-1,4,7,10-tetraacetic acid). J. Chem. Soc., Dalton Trans. 1998, (2), 243-248. 120. Lima, L. M. P.; Esteban-Gómez, D.; Delgado, R.; Platas-Iglesias, C.; Tripier, R., Monopicolinate Cyclen and Cyclam Derivatives for Stable Copper(II) Complexation. Inorg. Chem. 2012, 51 (12), 6916- 6927. 121. Rodriguez-Rodriguez, A.; Garda, Z.; Ruscsak, E.; Esteban-Gomez, D.; de Blas, A.; Rodriguez-Blas, T.; Lima, L. M. P.; Beyler, M.; Tripier, R.; Tircso, G.; Platas-Iglesias, C., Stable Mn2+, Cu2+ and Ln3+ complexes with cyclen-based ligands functionalized with picolinate pendant arms. Dalton Trans. 2015, 44 (11), 5017-5031. 122. Peisach, J.; Blumberg, W. E., Structural implications derived from the analysis of electron paramagnetic resonance spectra of natural and artificial copper proteins. Arch. Biochem. Biophys. 1974, 165 (2), 691-708. 123. Hancock, R. D.; Salim Shaikjee, M.; Dobson, S. M.; Boeyens, J. C. A., The Stereochemical activity or non-activity of the ‘Inert’ pair of electrons on lead(II) in relation to its complex stability and structural properties. Some considerations in ligand design. Inorg. Chim. Acta 1988, 154 (2), 229-238. 124. Kodama, M.; Kimura, E., A thermodynamic and kinetic interpretation of the macrocyclic effect. Polarographic studies on copper(II) 1,4,7,10-tetraazacyclododecane complexation. J. Chem. Soc., Chem. Commun. 1975, (9), 326-327. 125. Kodama, M.; Kimura, E., Equilibria of complex formation between several bivalent metal ions and macrocyclic tri- and penta-amines. Dalton Trans. 1978, (9), 1081-1085. 126. Motekaitis, R. J.; Rogers, B. E.; Reichert, D. E.; Martell, A. E.; Welch, M. J., Stability and Structure of Activated Macrocycles. Ligands with Biological Applications. Inorg. Chem. 1996, 35 (13), 3821-3827. 127. Alies, B.; Sasaki, I.; Proux, O.; Sayen, S.; Guillon, E.; Faller, P.; Hureau, C., Zn impacts Cu coordination to Amyloid-ß, the Alzheimer's peptide, but not the ROS production and the associated cell toxicity. Chem. Commun. 2013, 49 (12), 1214-1216. 128. Royal, G.; Dahaoui-Gindrey, V.; Dahaoui, S.; Tabard, A.; Guilard, R.; Pullumbi, P.; Lecomte, C., New Synthesis of trans-Disubstituted Cyclam Macrocycles – Elucidation of the Disubstitution Mechanism on the Basis of X-ray Data and Molecular Modeling. Eur. J. Org. Chem. 1998, 1998 (9), 1971- 1975. 129. Mato-Iglesias, M.; Roca-Sabio, A.; Pálinkás, Z.; Esteban-Gómez, D.; Platas-Iglesias, C.; Tóth, É.; de Blas, A.; Rodríguez-Blas, T., Lanthanide Complexes Based on a 1,7-Diaza-12-crown-4 Platform Containing Picolinate Pendants: A New Structural Entry for the Design of Magnetic Resonance Imaging Contrast Agents. Inorg. Chem. 2008, 47 (17), 7840-7851. 130. Oxford Diffraction Ltd, A., U.K. Crysalis software system, version 1.171.28 cycle 4 beta, 2005. 131. Sheldrick, G., A short history of SHELX. Acta Cryst. A 2008, 64 (1), 112-122.

~ 124 ~

Chapter II: Cu ion chelation

132. Glasoe, P. K.; Long, F. A., Use of glass electrodes to measure acidities in deuterium oxide. J. Phys. Chem. 1960, 64 (1), 188-190. 133. Rossotti, F. J. C.; Rossotti, H., Potentiometric titrations using Gran plots: A textbook omission. J. Chem. Educ. 1965, 42 (7), 375-378. 134. Schwarzenbach, G.; Flaschka, H. A., Complexometric titrations. Methuen: London, 1969. 135. Delgado, R.; do Carmo Figueira, M.; Quintino, S., Redox method for the determination of stability constants of some trivalent metal complexes. Talanta 1997, 45 (2), 451-462. 136. Gans, P.; Sabatini, A.; Vacca, A., Investigation of equilibria in solution. Determination of equilibrium constants with the HYPERQUAD suite of programs. Talanta 1996, 43 (10), 1739-1753. 137. Alderighi, L.; Gans, P.; Ienco, A.; Peters, D.; Sabatini, A.; Vacca, A., Hyperquad simulation and speciation (HySS): a utility program for the investigation of equilibria involving soluble and partially soluble species. Coord. Chem. Rev. 1999, 184 (1), 311-318.

~ 125 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

Chapter III. Impact of Zn(II) on the Cu(II) chelation

This chapter focuses on the impact of Zn(II) ion on the Cu(II) chelation within the Alzheimer’s disease context. This work is biologically relevant since Zn(II) is present in very high concentrations in the synaptic cleft compared to the Cu ions. Furthermore, Cu and Zn ions are found co-localised with the Aβ peptide in the senile plaques. That is the reason why it is biologically relevant to investigate the interplay between Cu and Zn ions with the Aβ peptide and the drug-candidates that are based on the chelation of Cu from Aβ.

A first part describes the state of the art regarding the impact of Zn(II) on the interaction between Cu ions and Aβ peptide. Metal ions coordination, ROS production and aggregation of the peptide are under focus. Moreover, the impact of Zn ions on the Cu ion chelation is reported and developed in the III-B section. A second part focuses on the Cu(II) chelation from a thermodynamic point of view. Studies regarding the removal of Cu(II) by different chelators from Aβ in the presence of Zn(II) have been performed by EPR, UV-Visible and XANES spectroscopies. The impact of this removal on the arrest of ROS production and on the aggregation have then been studied by fluorescence and UV-Visible spectroscopies. Aggregates have been imaged by AFM. The last part describes another proof of concept: the “pull-push” effect. For some ligands, Zn ions are necessary in order to remove Cu ions from Aβ. Zn(II) pulls Cu ions out from the peptide et pushes them into the chelator. This concept is described and experimentally illustrated by EPR, XANES and UV-Vis spectroscopies.

State of the art: the mutual interactions of Cu and Zn ions on the Aβ peptide

This part summarises the state of the art about the interactions of Zn(II) on the interplay between Cu ions and Aβ peptide. It is composed of a perspective article published in Dalton Transactions in 2017 and a summary of this perspective. I have co-written this review.

III-A.i Summary

This perspective summarises the different investigations published until now on the mutual interactions of Cu ions and Zn(II) with the Aβ peptide.

~ 126 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

In AD brains, a dyshomeostasis of Cu and Zn ions is reported: extracellular Cu(II) level is too high while the intracellular concentrations of Cu(I) are too low. Regarding Zn(II) concentrations, the tendency is not so clear. Indeed, some groups have found Zn ions in concentrations too high, and other ones, in concentrations too low. Furthermore, Cu and Zn ions have been found in the senile plaques and in vitro studies have shown that not only Cu ions but also Zn ions can interact with the Aβ peptide. These metal ions can also modulate the aggregation of Aβ, but again no tendency in vitro has been reached yet. Aggregation is important because oligomeric forms of Aβ are considered to be the most toxic species to cells. All the studies agree on the importance of the metal:peptide ratio and the fact that Cu(II) and Zn(II) ions both have an effect on the aggregation but that effect differs between Cu(II) and Zn(II). Another potential toxic mechanism is the ROS production catalysed by the Cu-Aβ complex, which can attack and degrade the surrounding biomolecules. That is why Cu(II) chelation is an important part of the Alzheimer’s therapeutic investigations since the last years. It has been reported that not only removing the Cu(II) from Aβ and redox silencing the Cu ions by the chelator might be important, but also the subsequent transport of Cu into the cell and its release as Cu(I) (such chelators are often called metallophores). Some of the chelators, such as clioquinol and then PBT2, have been under clinical trials but failed in phase II. Nowadays, one of the hypothesis for these failures is that the selectivity (i.e. the ratio between the affinity constant for Cu(II) and the one for Zn(II)) of the chelators between Cu(II) and Zn(II) is not high enough: the ligand removes Zn ions which then precludes removal of Cu ions. This could indicate that Cu(II) is a therapeutic target but that the interactions of Zn(II) (or other factors) has to be taken into account as well. In this perspective, the coordination of the homometallic (Cu(II) or Cu(I) or Zn(II)) and of the hetero-bimetallic (Zn(II) and Cu(II) or Cu(I)) complexes are reported, then the impact of Zn ions on the ROS production by Cu-Aβ and on the Cu-Aβ aggregation is described, and finally the impact of the presence of Zn(II) on the Cu ion removal by different chelators is discussed.

III.A.i.1.1 Coordination

The first part of this perspective reports the most accepted coordinations of the monometallic and hetero-bimetallic complexes (Figure III-1). The coordination sites of the monometallic complexes are described in the part I-B.iii.1. The coordination of the hetero-bimetallic complexes has been probed by different spectroscopic and potentiometric techniques. Regarding the Cu(II) coordination, Zn ions have an impact on it since Cu and Zn binding sites share some amino acids as ligands. Zn(II) coordination keeps the His6 as ligand, leading to the displacement of the Cu(II) binding site from component I to component II at physiological pH. The other ligands for the Zn(II) coordination are not yet been described, but the other His residue should be removed to this coordination at lower pH. Regarding

~ 127 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

Cu(I), Cu(I) keeps the same coordination than in the monometallic Cu(I)-Aβ and it imposes its coordination to the Zn ions, which loose the second His residue in its coordination. In brief, Cu(I/II) and Zn(II) have a mutual interference in their binding site to Aβ. It would be interesting to know the impact of this mutual interaction on the affinity constant of the metal ions for Aβ.

Figure III-1. Proposed models for the Cu(I), Cu(II) and Zn(II) coordination sites to Aβ as well as the hetero-bimetallic coordination..

III.A.i.1.2 ROS production and aggregation

The second part of this perspective focuses on the impact of Zn(II) on the aggregation and the ROS production by the complex Cu-Aβ. First of all, the aggregation of the Aβ peptide is modulated by Cu and Zn ions and these aggregations are highly dependent on the metal:peptide ratio, on the pH, on the temperature, etc. In the case of the Cu-Aβ aggregation, the metal:peptide ratio is very important. Indeed, in a sub-stoichiometric ratio of Cu ions compare to Aβ, fibrils are formed, whereas in at least a stoichiometric ratio, toxic oligomers are formed. Zn(II) impacts the Aβ aggregation even at a very low ratio, and at a stoichiometric ratio, fibrils as well as amorphous aggregates are formed. Note that the Zn(II)-induced fibrils are different from the apo-fibrils. The aggregation in the presence of both Cu(II) and Zn(II) has been reported to be the same than the aggregation with Zn ions only. Then, the ROS

~ 128 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation production by monomeric species in the presence of Zn(II) is reported. Indeed, there is a mutual interference of Cu and Zn ions in the coordination of the metal, in the aggregation of the peptide; it is biologically relevant to determine whether there is an impact of Zn(II) on the ROS production. Nevertheless, the tendency is not clear. Some studies show that Zn(II) has a protective effect towards the ROS production, increasing the survival of cells and decreasing the H2O2 production. This effect would be due to the displacement of the binding site of Cu(II) from component I to component II. Nevertheless, other studies confirm that Cu(II) is displaced at another site upon Zn(II)-binding, but they disagree on the fact that component II is redox silent. The presence of Zn(II) do not change the ROS production. This result is in line with a recent study explaining that the Cu(II) coordination is not the key parameter of the ROS production at all. Finally, the ROS production by aggregated species is reported in this perspective. One study describes that for a 0.5:1 Cu:Aβ ratio, the monomeric complex produces more ROS than the aggregate one. Another study shows that for a stoichiometric ratio, Cu- Aβ complex aggregates produce more ROS than the Cu,Zn-Aβ complex aggregates. In other words, the aggregated species should produce less ROS than the monomeric ones, and the presence of Zn should have a protective role on the ROS production by aggregates. In brief, nowadays, there are not enough studies with both Cu and Zn ions on the ROS production and/or the aggregation. No clear tendency can be reached with these few investigations, except that Zn(II) has not a dramatic effect on Cu-Aβ ROS production, for the monomeric species.

III.A.i.1.3 Cu(II) chelation

The last part of this perspective focuses on the impact of the presence of Zn(II) on the withdrawal of Cu(II) from Aβ by a chelator. The Cu(II) chelation in the AD context has been developed in the last years. A ligand needs a higher affinity constant for Cu(II) than the Aβ peptide, but its affinity constant has also to be lower than the one of essential metalloproteins, in order to not remove the essential Cu(II). As previously shown, Zn(II) impacts the Cu ion coordination, the ROS production and the aggregation. Therefore, it is an important parameter to take into account in the Cu chelation. The first studies regarding the impact of Zn(II) on the Cu chelation have been performed with the metallothionein Zn7-MT-3 (Figure III-2, left). There is a swap of metal ions between Cu(II) bound to Aβ and Zn(II) bound to Zn7-MT-3, removing Cu(II) from Aβ and forming an air stable Cu(I)4-thiolate cluster inside the metallothionein. This cluster silences the ROS production. Then the same results are shown with the MT-2A. Moreover, with the MT-2A, the same experiments with the apo-MT-2A are performed: Cu(II) stays bound to Aβ and the ROS production is not reduced. This sheds light on the importance of the presence of Zn(II). Then this kind of experiments is performed with synthetic ligands (Figure III-2, right). The two studied ligands, with a higher affinity constant for Cu(II) than the one of

~ 129 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

Aβ, are able to remove Cu(II) from Aβ. Nevertheless, in the presence of Zn(II), only one of them removes Cu(II) from the peptide. Indeed, they need a selectivity for Cu(II) towards Zn(II), i.e. a ratio between the affinity constant for Cu(II) and the one for Zn(II), and this is the case for both of them. They need also that this selectivity is higher than the selectivity of Aβ; if not they are not able to remove Cu(II) from Aβ. Note that this part is detailed in part III-B. Some future work or ideas to take into account are then given. All of these studies prove that the Cu chelation has to be performed in a Zn(II)- rich environment, even for Cu(I). Then, if Zn(II) can impact the Cu-chelation in the AD context, other biologically relevant ions such as Fe, Ca, Mg, etc. may also have an important role. Finally, as the thermodynamics are important in the Cu ion removal, the kinetics of removal of Cu ions from Aβ can also be an important parameter and should be studied next.

Figure III-2. Scheme representing the swap of metallic ions between metallothioneins or chelators and the Aβ peptide. The impact on the ROS production is also illustrated.

In this perspective, the mutual interactions of Cu and Zn ions with the Aβ peptide are described. Zn(II) has an important impact on the relationship between Cu ions and Aβ peptide. Zn(II) displaces Cu(II) from a binding site to another one. Moreover, it influences the ROS production of the Cu-Aβ complex. The aggregation of the peptide is also influenced by the presence of Zn(II): it appears that Zn(II) should impose its metal-induced aggregation. All of these studies shed light on the high impact of Zn(II) on the Cu-Aβ complex and its properties. Zn(II) also impacts the Cu chelation. In order to remove Cu ions from Aβ in the presence of Zn(II), the ligand needs not only a higher affinity constant for Cu than Aβ, but also a higher selectivity of Cu over Zn than the Aβ. This selectivity issue has to be taken into account also for the Cu(I) removal.

~ 130 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

III-A.ii Perspective

~ 131 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 132 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 133 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 134 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 135 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 136 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 137 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 138 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 139 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 140 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 141 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

The thermodynamic study

This section focuses on the thermodynamic aspect of the Cu(II) withdrawal from Cu(II)-Aβ by a chelator in the presence of Zn(II). It is composed of a summary of the article published in Dalton Transactions in 2016, the article itself and finally the supporting information. Note that this article is part of the review above.

III-B.i Summary

This article describes the impact of Zn(II) ions on the Cu(II) chelation from Aβ. The first step of the chelation therapy, meaning the removal of Cu ions from Aβ in AD brains, was to understand how a ligand can sequester Cu(II) ion from the Aβ peptide and how it can stop the ROS production. Many studies have proposed organic as well as peptidic ligands in order to remove Cu(II) from the Aβ peptide. But, as previously explained, Zn(II) ion is present in high concentrations in the synaptic cleft, and in the senile plaques. Some works have shown that Zn(II) has an influence on the Cu(II) coordination with Aβ, on the ROS production by the Cu(II)-Aβ complex, and on the metal-induced aggregation of the peptide. Thus, the second step of the chelation therapy studies before going to more complex systems like studies in cellulo, has been the understanding of the 4-partner system, namely Cu(II), Zn(II), Aβ and a chelator. This four-partner system has two metal ions and two ligands: it is important to know if the chelator can remove Cu ions from Aβ in the presence of Zn ions or if it will chelate Zn ions. The following equilibrium is considered:

Cu(II)-Aβ + Zn(II) + L Zn(II)-Aβ + Cu(II)-L

In order to determine if this equilibrium is on the right or on the left, spectroscopic techniques are used to distinguish if Zn(II) and/or Cu(II) is bound to Aβ or to the chelator.

A good chelator needs a higher affinity constant for Cu(II) than Aβ: the metal ion is chelated by the ligand. Nevertheless, in the presence of Zn(II) ion, this parameter is not enough for the removal of Cu(II) from Aβ. In this article, two Cu(II) chelators, L2 and Lc, have been studied: both have an affinity constant for Cu(II) higher than Aβ and they both sequester the metal ion in the absence of Zn(II).

The study of the swap of metal ions, Cu(II) and Zn(II), between these two chelators and Aβ is the first part of this work (see equilibrium above). UV-Visible, EPR and XANES are used in order to perform this experiment. For the L2 ligand, all the results lead towards the swap of metal ions (equilibrium on the right), whereas not for the Lc ligand (equilibrium on the left). For the latter, in the presence of

~ 142 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

Zn(II), Cu(II) stays bound to Aβ: Lc is not able to remove Cu(II) from the peptide whereas L2 can, altough both have a higher affinity constant for Cu(II) than Aβ. A thermodynamic explanation is proposed for this phenomenon. The main difference between these two chelators is their selectivity. The selectivity of a ligand is the ratio between its affinity for Cu(II) and its affinity for Zn(II). Moreover, a high affinity constant for Cu(II) does not mean that the corresponding selectivity is high, and vice versa. In order to remove Cu(II) from Aβ in the presence of Zn(II), the chelator needs not only a higher affinity constant for Cu(II) than Aβ, but also a higher selectivity, which is already relatively high for Aβ (about 104.2 at pH 7.1). The Cu(II) over Zn selectivities of L2 and Lc are about 107.7 and 102.0, respectively, at pH 7.1. Thermodynamically, only L2 is able to sequester Cu(II) in the presence of Zn(II).

A second part of the study is dedicated to the efficiency of these two chelators in preventing the ROS production. This work is performed via two different experiments: the first one consists in the following of the kinetic of ascorbate consumption mirroring the ROS production and the second one is the kinetic of formation of 7-OH-CCA, a fluorescent compound formed via the reaction between the coumarin carboxylic acid (CCA) and the HO•. Without Zn(II), both ligands stop the formation of these toxic species, at least reduce considerably their production, meaning that when Cu(II) is bound to L2 or Lc, the associated complexes do not produce ROS. Then, in order to mimic the high concentrations of Zn(II) in the synaptic cleft, Zn(II) is preloaded to the ligands: in a rich Zn(II)-environment, the probability of the ligand to not chelate first Zn(II) is very low. When Zn(II) is preloaded to the L2 ligand, there is more or less no effect of Zn(II) on the ROS production, contrary to the Lc ligand. In the presence of Zn(II), Lc is not able to stop the ROS production anymore. This result is relatively obvious since in the presence of Zn(II), Cu(II) is chelated by the Aβ peptide and not by Lc (equilibrium above on the left), and so, the Cu-Aβ complex can produce ROS.

The last study of this work focuses on the aggregation of the Aβ peptide with the L2 ligand. Note that the aggregation with Lc has not been performed because in the presence of Zn(II), Cu stays bound to Aβ and the aggregation would be the same than without the ligand. The aggregation of Aβ is followed by the ThT fluorescence, and the samples have then been visualized by AFM. The hypothesis of this thesis is that apo-Aβ and Zn(II)-Aβ aggregate into fibrils, whereas Cu(II)-Aβ aggregates into oligomers or protofibrils. If the L2 ligand is added to Cu(II)-Aβ before the starting of the aggregation, L2 chelates Cu(II) and Aβ peptide can aggregate into fibrils, as an apo-peptide. The same result is obtained if Zn(II)-L2 is added to Cu(II)-Aβ. This confirms that L2 is able to withdraw Cu(II) from Aβ even when preloaded with Zn(II).

~ 143 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

This article proposes a thermodynamic criterion for a Cu(II) chelator in a rich Zn(II)-environment: the selectivity of Cu(II) over Zn ions. It has to be higher than Aβ, already important (about 104.2 at pH 7.1), due to the different binding site of Cu(II) and Zn(II). When a ligand satisfies this criterion, Cu(II) can be withdrawn from Aβ even in the presence of Zn(II). This leads to a stop of the ROS production (in case of Cu(II)-ligand does not produce ROS itself) and to a fibrillary Zn(II)-induced aggregation, aggregates which are proposed to be less toxic than oligomers. Figure III-3 illustrates the thermodynamic issue described in this part.

Figure III-3. Scheme of the possible approach on the chelatotherapy describes in this part. While without Zn(II), L2 and Lc are able to remove Cu(II) from Aβ and stop the ROS production, in the presence of Zn(II) only L2 is able to remove Cu(II) and stop the ROS production. This is due to the fact that L2 and Lc have both an affinity constant for Cu(II) higher than the one of Aβ, but only L2 has a Cu(II) over Zn(II) selectivity higher than the one of Aβ.

This study is a first step before a more elaborate system. Indeed, it would be interesting to investigate the impact of the surrounding biomolecules within the synaptic cleft and also other metal ions on the Cu(II) chelation. Moreover, this parameter can be applied to the Cu(I) therapy, but also to other amyloidogenic diseases such as Parkinson or Amyotrophic Lateral Sclerosis, in which a dyshomeostasis of metal ions is involved.

~ 144 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

III-B.ii Article

~ 145 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 146 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 147 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 148 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 149 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 150 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 151 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 152 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

III-B.iii Supporting information

How Zn can impede Cu detoxification by chelating agents in Alzheimer's Disease: a proof-of-concept study

Amandine Conte-Daban,[a] Adam Day,[a] Peter Faller[a,b] and Christelle Hureau[a]

[a] A. Conte-Daban, A. Day, Prof. P. Faller, Dr. C. Hureau

CNRS, LCC (Laboratoire de Chimie de Coordination)

205 route de Narbonne, BP 44099, 31077 Toulouse Cedex 4 (France) and University of Toulouse, UPS, INPT, 31077 Toulouse Cedex 4 (France)

E-mail: [email protected]

[b] Prof. Dr. P. Faller

present address : Institute de Chimie (UMR 7177), 4 rue B. Pascal, F-67000 Strasbourg, France

SUPPORTING INFORMATION

Materials and methods

Cu(II) and Zn(II) coordination sites to Aβ Ligands ROS detection methods Equations Apparent affinity and selectivity values of the chelators

UV-Vis, EPR and XANES signatures of Cu removal from Aβ ROS production assays Aggregation assay References

~ 153 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 154 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 155 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 156 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 157 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 158 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 159 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 160 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 161 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 162 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 163 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 164 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

~ 165 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

The Cu(II) “pull-push” effect

This part focuses on a new concept: the “pull-push”. It is composed on a theoretical explanation of the concept, an experimental section and the results obtained with three ligands.

III-C.i Theoretical concept

As previously explained, chelatotherapy is one of the therapeutic approach against AD. It exists different kind of chelators (for more details, see II-A.). Some of them have an affinity constant for Cu ions too high, meaning that they could remove essential Cu ions from other metalloproteins. Other chelators have a too low affinity constant for Cu ions; therefore they are likely not efficient against AD since Cu ions stay bound to the Aβ peptide. Lastly, some chelators have an affinity constant for Cu ions in the range of the one of the Aβ peptide. Thus, there is an equilibrium between Cu ions bound to Aβ and Cu ions bound to the chelator.

The importance of Zn ions on the Cu chelation has been demonstrated in III-B. Chelators with a higher affinity for Cu(II) than Aβ such as Lc are not yet able to remove Cu ions from Aβ in the presence of Zn ions. The thermodynamic explanation is the selectivity of Cu over Zn ions: to be an efficient Cu(II)- chelator in the presence of Zn ions, a selectivity of Cu(II) ions over Zn ions higher than the one for Aβ, which is of 4 orders of magnitude at pH 7.1, is needed (Table III-1, L >> Aβ). The following equilibria illustrate this phenomenon.

Equilibrium 1:

↔ , with .

Equilibrium 2:

↔ , with

.

Equilibrium 3:

↔ , with

~ 166 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

. .

Equilibrium 4:

↔ , with

. . . .

Equilibria 1 and 2 reflect the affinity constants of Aβ peptide or a ligand for a metal ion M, such as Cu or Zn ions. Equilibrium 3 is the equilibrium for the Cu removal from Aβ by a chelator L. Equilibrium

4 is the equilibrium for the Cu removal in the presence of Zn ions. K1 and K2 are the associated equilibrium constants. The following Table III-1 shows some examples of the impact of the selectivity of the chelator on the equilibrium 4. Note that the affinity constants of Aβ for Cu(II) and Zn(II) at pH 7.1 are approximately 109 and 105 M-1 respectively. The ideal chelator must have a selectivity at least equal to 7 order of magnitude in order to remove all Cu ions from Aβ in the presence of Zn ions. Note that the hypothesis is that all Cu ions are removed from Aβ when it stay 1/1000 of Cu ions bound to Aβ; the equilibrium constant being 103.

Table III-1. Table illustrating different cases for the equilibria 3 and 4. The percentages of Cu bound to L are also reported.

Cu(II) affinity K % Cu(II)-L Selectivity of L K % Cu(II)-L constant of L 1 2 101 10-3 3 % L >> Aβ 102 90 % 104 1 50 % 1011 >> 109 107 103 97 % 101 10-3 3 % L = Aβ 1 50 % 104 1 50 % 109 = 109 107 103 97 % 101 10-3 3 % L << Aβ 10-2 10 % 104 1 50 % 107 << 109 107 103 97 %

The “pull-push” concept takes into account the fact that the addition of Zn ions can pull Cu ions out of the Aβ peptide and push it into the chelator. The chelator becomes more efficient in the removal of Cu ions in the presence of Zn ions than without. There are two main approaches for the “pull-push”.

~ 167 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

The first one applies to chelators with an affinity constant close to the one for Aβ (109 M-1 at pH 7.1, see Table III-1, L = Aβ). This means that half of Cu ions is bound to Aβ peptide without Zn ions. Ideally, the addition of Zn ions should move the equilibrium 4 from about 50 % to near 100 % of Cu-L. This is

the case if the selectivity of the chelator is at least 7 orders of magnitude (see Table III-1, L = Aβ); -1 -1 10 M and 10 M at pH 7.1. The second approach applies to chelators with a very low affinity constant for Cu(II) (Table III-1, , L << Aβ). Note that if the affinity constant of the chelator for Cu ions is low, the selectivity cannot be equal to 7 orders of magnitude because the affinity constant for Zn ions is already very low and this is difficult to reach such affinities for only one molecule (the case of an affinity for Cu(II) at 107 and a selectivity at 107 is probably impossible to reach). In this second case, the addition of Zn(II) should move the equilibrium from about 10 % to 50 % of Cu(II) bound to L. The first approach is illustrated with three ligands and the experimental part is described in the following section.

III-C.ii Experimental section

Chemicals. Reagents were commercially available and were used as received. All the solutions were prepared in milliQ water (resistance: 18.2 MΩ.cm).

The Cu(II) ion source was CuSO4.5H2O, bought from Sigma-Aldrich. A stock solution was prepared at 25 mM. The Zn(II) ion source was ZnSO4.H2O, bought from Sigma-Aldrich. A stock solution was prepared at 100 mM.

HEPES buffer (sodium salt of 2-[4-(2-hydroxyethyl)piperazin-1-yl]ethanesulfonic acid) was bought from Sigma-Aldrich. A stock solution was prepared at 500 mM, pH = 7.1.

Ascorbate was bought from Sigma-Aldrich. A 5 mM stock solution was prepared in Milli-Q water just before beginning the experiments. Because ascorbate can be oxidized by air, a fresh solution was prepared every day.

Ligand and peptides. The 3,4-bis(oxamato)benzoic acid ligand L (see Figure III-2) was obtained from a collaboration with L. Lisnard (University Pierre et Marie Curie, Paris). For the synthesis, see ref.1 A 8.5 mM stock solution was prepared, increasing the pH until solubilization, determined by UV-Vis titration with a titrated Cu(II) solution following the absorption band at 330 nm of the complex according to ref.2.

Aβ16 (DAEFRHDSGYEVHHQK) and Aβ28 (DAEFRHDSGYEVHHQKLVFFAEDVGSNK) were bought from GeneCust (Dudelange, Luxembourg) with purity grade > 95 %. Stock solutions of the peptides

~ 168 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation were prepared by dissolving powder in milliQ water (resulting pH ~ 2) and were titrated by UV-Vis

-1 -1 3 absorption of Tyr10 considered as free tyrosine (at pH ~ 2.0, ε276 - ε296 = 1410 cm .M ).

ABH and BAH where B corresponds to the β-alanine amino acid were bought from Protéogenix (Strasbourg, France). The stock solutions were prepared by solubilizing the peptides in milliQ water and were titrated by UV-Vis with a titrated Cu(II) solution following the d-d band absorption of the complex at 545 nm and 536 nm respectively,4 according to ref. 2

Methods.

Electron Paramagnetic Resonance. Electron Paramagnetic Resonance (EPR) data were recorded using an Elexsys E 500 Bruker spectrometer, operating at a microwave frequency of approximately 9.5 GHz. Spectra were recorded using a microwave power of 5 mW across a sweep width of 120 mT (centred at 310 mT) with modulation amplitude of 0.5 mT. Experiments were carried out at 120 K using a liquid nitrogen cryostat.

EPR samples were prepared from stock solution of ligand diluted down to 0.2 mM in H2O. 190 µM

65 65 65 of Cu(II) were added from 25 mM Cu(NO3)2 home-made stock solution from a Cu foil. Samples were frozen in quartz tube after addition of 10% glycerol as a cryoprotectant and stored in liquid nitrogen until used.

X-ray Absorption Near–Edge Structure (XANES) spectra were recorded at the BM30B (FAME) beamline at the European Synchrotron Radiation Facility (ESRF, Grenoble, France).5 The storage ring was operated in 7/8+1 mode at 6 GeV with a 200 mA current. The beam energy was selected using a

Si(220) N2 cryo-cooled double-crystal monochromator with an experimental resolution close to that theoretically predicted (namely ~ 0.5 eV FWHM at the Cu energy). The beam spot on the sample was approximately 300 x 100 µm2 (H x V, FWHM). Because of the low Cu(II) concentration, spectra were recorded in fluorescence mode with a 30-element solid state Ge detector (Canberra) in frozen liquid cells in a He cryostat. The temperature was kept at 20 K during data collection. The energy was calibrated with Cu metallic foil, such that the maximum of the first derivative was set at 8979 eV. At least three scans recorded on different spots were averaged. The samples were prepared from stock solutions of ligand and Cu(II) diluted down to approx. 1.0 mM or 200 µM in buffered solution. Samples were frozen in the sample holder after addition of 10% glycerol as a cryoprotectant and stored in liquid nitrogen until used. Cu(II) photoreduction was avoided by changing the position of the beam between each scan. It was considered that during the first 20 minutes of recording (more or less one scan) the photoreduction is insignificant.

~ 169 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

UV-Visible spectrophotometry. UV-vis spectra were recorded on a spectrophotometer Agilent 8453 at 25°C in 1 cm path length quartz cuvette, with stirring at 800 rpm.

Experiment illustrating the impact of Zn(II) on the Cu(II) removal from Aβ by the ligand L. The experiments have been monitored by UV-Vis in HEPES buffer 0.1 M at a resulting pH = 7.1. Aβ and Cu(II) (50 or 20 µM) were mixed and the ligand L (50 or 20 µM) was added. When the equilibrium is reached, successive additions of Zn(II) (0 to 20 equiv. in total) were added. Each addition of Zn(II) is made once the thermodynamic equilibrium of the previous reaction is reached.

ROS production experiment followed by Ascorbate consumption. The experiments have been monitored by UV-Vis in HEPES buffer 0.1 M at a resulting pH = 7.1. Ascorbate is diluted down 100 µM, the peptide and the ligands are at 12 µM and Cu(II) at 10 µM.

III-C.iii Illustration of the “pull-push” concept

To illustrate the “pull-push” concept, three chelators were used. Their structures are shown in Figure III-4. Note that L* stands for one of these ligands. The Cu(II) affinity constants of L, ABH and BAH are 3.2 109 M-1, 3.0 108 M-1 and 3.7 109 M-1 at pH 7.1, respectively.6 Note that these affinity constants are in the same range than the one of Aβ which is 1.6 109 M-1.1, 7 The calculations of the Equilibrium 3 give the K1 values of 2.0, 0.2 and 2.3, respectively. These values indicate that in the absence of Zn(II), L and BAH removes ~ 60 % of Cu(II) from Aβ and ABH removes only 30 % of Cu(II). Note that the affinity constants for Zn ions of these three ligands is not known (but should be very low based on the structures of the ligands). The predictive calculations of Equilibrium 4 are not possible without these values. During this thesis, different experiments have been performed in order to know if these ligands are able to remove Cu(II) from Aβ, if the removal is in good agreement with the theory and what is the behavior of these ligands in the presence of Zn(II). First, EPR, then XANES and finally UV-Vis experiments are described.

Figure III-4. Structures of the three ligands used in the “pull-push” concept. L is a synthetic ligand, ABH and BAH are synthetic peptides. B stands for the β-alanine.

~ 170 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

III-C.iii.1 EPR experiments

Several EPR experiments have been performed. Figure III-5 shows one of them. A competition between Aβ and the ligands with or without Zn ions were carried out. For L, ABH and BAH, there is a good agreement with the thermodynamic calculations. Indeed, linear combinations of the EPR experiments shown in the Figure III-5 reveales that L removes 60 % of Cu(II) from Aβ, ABH 25 % and BAH 50 % (Table III-2). When Zn ions are also in competition, the percentages of the Cu(II) removal change. Using the calculations of the Equilibrium 4 and linear combinations of the EPR experiments (Figure III-5), the percentage in the presence of Zn(II) are 80 % of Cu(II) removed by L from Aβ, 45 % by ABH and 70 % by BAH (Table III-2). These experiments illustrate the “pull-push” effect. Note that there is a kinetic effect in the Cu(II) chelation. At t0, meaning without incubation, ABH and BAH remove only

10 % of Cu(II) from Aβ, not consistent with the thermodynamic calculations. After t1 (64 h in the fridge) ot t2 (7h at room temperature), the percentages of Cu(II) removed from Aβ are consistent with the thermodynamic caluclations (Table III-2). Furthermore, the percentages obtained for Equilibria 3 and 4 change between the different experiments, but stay by less than 10 %.

Figure III-5. EPR experiments of competition between Aβ and L (Panel A), ABH (Panel B) or BAH (Panel C). (a) Aβ + Cu(II), (b) L* + Cu(II), (c) Aβ + Cu(II) + L*, (d) Aβ + Cu(II) + Zn(II)-L*. (L* = L, ABH or BAH). [L*] = [Aβ ] = [Zn(II)] = 200 μM, [65Cu(II)] = 190 μM, [HEPES] = 50 mM. pH = 7.1. T = 110 K. 10 % of glycerol was used as gryoprotectant. The incubation time for experiments for L and ABH is 64 h in the fridge, 7 h at room temperature for BAH.

Table III-2. Table reporting the percentages of Cu(II) removed by the ligands from Aβ without and with Zn(II).

Without Zn(II) With Zn(II) L 60 % 80 % ABH 25 % 45 %

BAH 50 % 70 %

~ 171 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

III-C.iii.2 XANES experiments

XANES is a very appropriate technique to probe the environment of Zn ions since they are silent in most of the classical spectroscopic techniques. XANES allows to follow Cu and Zn ion chelation in the same time and with the very same sample. Therefore, in order to illustrate the “pull-push” effect, XANES appears to be a very well appropriate technique.

Different experiments have been performed. The first two ones have been monitored with ligand, metal ions and the Aβ28 peptide at approximately 1 mM, minimal concentration required for our XANES conditions. Note that Aβ16 has not been used in order to limit the precipitation of Zn(II)-Aβ. The same experiments than for the EPR are carried out: Aβ + Cu(II) + L* and Aβ + Cu(II) + Zn(II)-L*. In the presence of Zn(II), even if Aβ28 was used, there was an issue with the precipitation. The “pull- push” effect was visible for the three ligands, leading to more or less 100 % of Cu(II) bound to the ligand in the presence of Zn(II). The “pull-push” effect worked well, maybe too well and was not consistent with the EPR experiments. The precipitation of Zn(II)-Aβ displaces the equilibrium through the formation of Zn(II)-Aβ (see Equilibrium 5 below). This implicates that the equilibrium goes towards the chelation of Cu(II) by L*. This is the reason why Cu ions are removed from Aβ by L* but this is not due to the “pull-push” effect only.

Equilibrium 5: Cu(II)-Aβ + Zn(II) + L* Zn(II)-Aβ + Cu(II)-L*

The other experiment was performed with Aβ16 at 200 μM; a concentration at which Zn(II)-Aβ does not precipitate. Nevertheless, 200 μM being a too low concentration for the beamline and detectors used; 1 mM is more or less the lower limit. To overcome this problem, 12 scans were accumulated instead of 3 when the concentration was around 1 mM. However, the spectra obtained were not exploitable. The noise of the spectra is too important, precluding any kind of calculations. In the ESRF at Grenoble where XANES was carried out, a new beamline for highly diluted samples has just opened. It could be interesting to perform these competition experiments in this beamline. However, all the current XANES results are useless.

III-C.iii.3 UV-Vis experiment for L

As previously described in ref.1, Cu(II)-L exhibits an important absorbance at 330 nm, ε = 18 000cm--1.M-1. Therefore, the “pull-push” effect can be follow by the detection of Cu(II)-L. This is a preliminary experiment. First, Aβ is complexed with Cu(II). Then, one equivalent of L is added. When

~ 172 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation the equilibrium is reached, one by one equivalent of Zn(II) is added, until 5 equivalents. Figure III-6 shows the kinetic of the formation of Cu(II)-L at 330 nm. Table III-3 exhibits the absorbance and the experimental percentages of Cu(II) bound to L, as a function of the number of equivalents of Zn(II). Without Zn(II), the percentage of Cu(II) bound to Aβ is quite high compare to what is expected based on the Ka values (78 % vs. 60 %). This may be due to a too high concentration of L in the cuve. The addition of each equivalent of Zn(II) until 5 equivalents increases the quantity of Cu(II) bound to Aβ until 100 %. For L, 4 or 5 equivalents of Zn(II) are needed to push all Cu(II) in the chelator.

Figure III-6. Kinetic of formation of Cu(II)-L, followed by UV-Vis spectroscopy at 330 nm. [Aβ] = [L] = [Cu(II)] = 20 µM, Zn(II) is added equivalent by equivalent. [HEPES] = 100 mM, pH = 7.1, 25°C.

Table III-3. Table bringing together the values of absorbance of Cu(II)-L as a function of the number of added equivalent of Zn(II). The percentages of Cu(II) bound to L compared to the total Cu(II) are given.

Number of 0 1 2 3 4 5 equivalents Abs (330 nm) 0,28 0,30 0,32 0,34 0,35 0,36 % Cu(II)-L 78 % 83 % 89 % 94 % 97 % 100 %

This experiment is not possible for the other two ligands ABH and BAH because they do not exhibit an important absorbance when bound to Cu(II). Nevertheless, a fluorescence experiment could be carried out. Indeed, it is possible to add an extrinsic fluorophore to the ligand (strategy A) or to use the fluorescence of the tyrosine residue of Aβ (strategy B)8 and, as the paramagnetism of Cu(II) quenches the fluorescence, it is possible to determine if Cu(II) is bound to Aβ or to the ligand. For the strategy A, if Cu(II) is bound to the ligand, there will be exctinction of the fluorescence contrary to if Cu(II) is bound to Aβ (detection of the fluorescence from the ligand). For the strategy B, this is the opposite.

~ 173 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

III-C.iii.4 ROS production experiments

As previously explained, the ligands have to stop the ROS production. To monitor this, Ascorbate consumption assay has been performed. The results are shown in the Figure III-7. Cu(II)-L* produces ROS. Different kind of experiments have been monitored: a pre-incubation of Cu(II) and L* before adding Ascorbate, use of 10 equivalents of L* to preclude the free Cu (the used concentrations are very low), but Cu(II)-L* produces ROS. Up to now, it is very difficult to understand why these complexes catalyse the ROS production. Note that the “pull-push” concept cannot be seen with the ROS production experiments since Cu(II)-L* doesn’t stop the ROS production.

Figure III-7. ROS production followed by the consumption of Ascorbate by UV-Vis spectroscopy, at 265 nm with a background correction at 800 nm. (a) L + Cu(II) + Asc, (b) ABH + Cu(II) + Asc, (c) BAH + Cu(II) + Asc. [L*] = 12 μM, [Cu(II)] = 10 μM, [Asc] = 100 μM, [HEPES] = 100 mM, pH = 7.1, T = 25°C.

III-C.iv Conclusion

In this part, the “pull-push” concept is described (Figure III-8). It consists in pulling Cu(II) out from the Aβ by the presence of Zn(II) and pushing Cu(II) into the chelator. The ligands able to do the “pull- push” have an affinity constant for Cu(II) in the same range or below than the one of Aβ. Their selectivity for Cu over Zn ions has to be higher than the one of Aβ in order to remove almost all Cu(II) from Aβ in the presence of Zn(II). Three ligands (L* = L, ABH and BAH) have been probed for the “pull- push” concept by EPR, XANES and UV-Vis spectroscopies. One equivalent of Zn(II) increases the quantity of Cu(II) removed from Aβ and chelated by L*. In the case of L, 5 equivalents of Zn(II) are needed to remove all Cu(II) from Aβ. For the other two ligands, fluorescence experiments would have to be carried out in order to know the number of equivalents of Zn(II) to remove «all» Cu(II) from Aβ. Note that it is quite difficult to monitor the “pull-push” experiments due to precipitation issues of Zn(II)-Aβ.

~ 174 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

Figure III-8. Scheme representing the “pull-push” concept, with the three ligands studied here.

L, ABH and BAH are good “pull-push” ligands. The problem is that they are not able to stop the ROS production; but they illustrate quite well the “pull-push” concept. It would be interesting to determine their affinity constant for Zn(II). Indeed, for example, in case of L, 5 equivalents and not only one are needed to remove all Cu(II) from Aβ, the Cu over Zn ion selectivity of L should be too low. Nevertheless, Zn(II) concentration in the synaptic cleft is 10 to 100 times higher than Cu ions, the need of 5 equivalents might not be a problem in vivo.

Note that the “pull-push” concept is a thermodynamic concept. It does not take into account kinetic issues. Furthermore, it is difficult to understand why the Cu(II)-L* complexes are not able to stop the ROS production. Maybe the conditions needed to fulfill the criteria of the affinity and selectivity for a “pull-push” ligand preclude it to stop the ROS production (geometry and kinetic issues?). Studying ligands with a very low affinity constant for Cu(II) to illustrate the second approach of the “pull-push” concept would be also interesting.

~ 175 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

Conclusion

This part focuses on the impact of Zn ions on the Cu(II) chelatotherapy against AD. It has been proposed that depending on the Cu(II) affinity constant and on the Cu(II) over Zn selectivity of the ligand, Zn ions can preclude the Cu(II) removal from Aβ or can trigger it. Both of these parameters are very important in the design of ligands against AD.

Maybe the better strategy would be the pro-drug one described with the “pull-push” concept. Indeed, for this strategy, the ligands have a low Cu(II) affinity constant, meaning that they could not be able to remove Cu(II) in vivo from any protein. Nevertheless, with a Cu(II) over Zn(II) selectivity around 105-106 (or higher if possible), the ligand is able to remove Cu(II) from Aβ only in the presence of Zn ions, as for example in the Zn(II)-rich synaptic cleft.

Zn ions are ejected from the neuron to the synaptic cleft and are re-taken very rapidly, meaning that Zn levels fluctuate constantly up and down. One can wonder if the important parameter of the kinetic of Cu(II) removal (detailed in the section II-2) is also important in the presence of Zn ions. It seems that it is always very important. The ligand needs a very important selectivity of Cu(II) over Zn to not chelate Zn ions during the ejection of these pools of Zn(II). Furthermore, the ligand needs to chelate Cu(II) very fast, as not to chelate Zn ions instead of Cu(II) ions. Some preliminary experiments have been performed, showing that, in vitro, even with a fast Cu(II) chelation, a ligand can chelate first Zn ions, leaving Cu(II) ions bound to Aβ. More studies are needed to understand how this issue could be resolved.

Another important parameter to take into account is the impact of other ions such as Ca or Fe ions or other biomolecules on the Cu(II) chelation. Indeed, Zn has been studied first because it is present in very high concentrations in the synaptic cleft. Nevertheless, as the impact of Zn is not insignificant, other molecules could also have an effect and it could be of interest to study the importance of this.

~ 176 ~

Chapter III: Impact of Zn(II) on the Cu(II) chelation

References

1. Conte-Daban, A.; Borghesani, V.; Sayen, S.; Guillon, E.; Journaux, Y.; Gontard, G.; Lisnard, L.; Hureau, C., Link between Affinity and Cu(II) Binding Sites to Amyloid-beta Peptides Evaluated by a New Water-Soluble UV-Visible Ratiometric Dye with a Moderate Cu(II) Affinity. Anal. Chem. 2017, 89 (3), 2155-2162. 2. Hureau, C.; Eury, H.; Guillot, R.; Bijani, C.; Sayen, S.; Solari, P. L.; Guillon, E.; Faller, P.; Dorlet, P., X-Ray and solution structures of CuGHK and CuDAHK complexes. Influence on their redox properties. Chem. Eur. J. 2011, 17 (36), 10151-10160. 3. Faller, P.; Hureau, C.; Dorlet, P.; Hellwig, P.; Coppel, Y.; Collin, F.; Alies, B., Methods and techniques to study the bioinorganic chemistry of metal-peptide complexes linked to neurodegenerative diseases. Coord. Chem. Rev. 2012, 256 (19-20), 2381-2396. 4. Nagaj, J.; Stokowa-Sołtys, K.; Zawisza, I.; Jeżowska-Bojczuk, M.; Bonna, A.; Bal, W., Selective control of Cu(II) complex stability in histidine peptides by β-alanine. J. Inorg. Biochem. 2013, 119, 85- 89. 5. Proux, O.; Biquard, X.; Lahera, E.; Menthonnex, J. J.; Prat, A.; Ulrich, O.; Soldo, Y.; Trévisson, P.; Kapoujvan, G.; Perroux, G.; Taunier, P.; Grand, D.; Jeantet, P.; Deleglise, M.; Roux, J.-P.; Hazemann, J.- L., FAME: A new beamline for X-ray absorption investigations of very-diluted systems of environmental, material and biological interests. Phys. Scr. 2005, 115, 970-973. 6. Nagaj, J.; Stokowa-Sołtys, K.; Zawisza, I.; Jeżowska-Bojczuk, M.; Bonna, A.; Bal, W., Selective control of Cu(II) complex stability in histidine peptides by β-alanine. J. Inorg. Biochem. 2013, 119, 85– 89. 7. Kowalik-Jankowska, T.; Ruta, M.; Wiśniewska, K.; Łankiewicz, L., Coordination abilities of the 1–16 and 1–28 fragments of β-amyloid peptide towards copper(II) ions: a combined potentiometric and spectroscopic study. J. Inorg. Biochem. 2003, 95 (4), 270-282. 8. Alies, B.; Renaglia, E.; Rózga, M.; Bal, W.; Faller, P.; Hureau, C., Cu(II) Affinity for the Alzheimer’s Peptide: Tyrosine Fluorescence Studies Revisited. Anal. Chem. 2013, 85 (3), 1501-1508.

~ 177 ~

General conclusion

General conclusion

Along this thesis, different proofs of concept regarding the Cu chelatotherapy against AD are described. The studies are split into two main parts: on one hand, the Cu chelation by a ligand from the Cu-Aβ complex and on the other hand the impact of Zn ions on this Cu chelation. Several questions were outlined in the introduction and the answers are given below.

Figure 1. Summary scheme of the different ligands studied in this thesis. Note that the blue part (PTA) is described in Annexe 1 and focuses on a Cu(I) chelator. Solid arrows describe a quasi-total Cu ion removal by the ligand from Aβ. A dotted line describes an equilibrium. An arrow with special arrow head describes a fast kinetic in the Cu removal from Aβ by a ligand.

The first proof of concept detailed in this thesis is about the possible kinetic issue in the Cu chelation (Figure 1, pink). The first question was: - Does the kinetic of Cu chelation by a ligand have an impact on the ROS production catalysed by Cu-Aβ complex? To address this question, cyclen and cyclam macrocyclic ligands have been studied. The production of ROS has been monitored. Indeed, they can stop the ROS production only if they are pre-incubated with Cu(II)-Aβ (allowing formation of the Cu(II)-complex with the macrocyclic ligands) before the addition of the reductant since the Cu chelation by these ligands is too slow. Thus, this demonstrates the impact of the Cu(II) chelation kinetic on the arrest of the ROS production by Cu-Aβ. - How is it possible to avoid that the ligands fail in the removal of Cu ions due to kinetic reasons (despite thermodynamically favoured) when designing a ligand against AD?

~ 178 ~

General conclusion

One of the possible answers is the addition of chelating arms. Indeed, cyclen and cyclam have been substituted with one or two picolinate arms. The following of the ROS production has also been performed with the substituted ligands. They are able to chelate Cu(II), even when it is found redox cycling during the ROS production, meaning that even in the presence of a mixture of Cu(I)/Cu(II), the substituted ligands are able to remove Cu from Aβ and stop the ROS production associated. In this case, there is no more need of incubation. Note that in the present case, the addition of chelating arms increases the Cu(II) chelation kinetic. Nevertheless, this strategy of adding chelating arms should not be efficient for all the ligands with Cu(II) chelation kinetic issues. Indeed, the addition of arms on linear ligands or macrocyclic ligands with another kind of cavity for example should not change the kinetics. - Does the complex geometry have an impact on the kinetic of the Cu chelation? Based on the results, as cyclam and cyclen complexes do not have the same geometry and present different behaviours regarding the Cu(II) removal from Aβ, the geometry of the complex could have an impact on the kinetic of the Cu chelation. Indeed, the Cu ion is found outside the cavity of the macrocycle for cyclen whereas it is inside the cavity for the cyclam. It should take more time to enter the cavity than to stay upon the cavity. Indeed, the Cu chelation by the cyclam is slower than for the cyclen. We have been able to, by this first proof of concept, describe the importance of the kinetic on the Cu chelation approach.

The second proof of concept on the Cu chelatotherapy is about the choice of the target (Figure 1, green). Indeed, Cu is the target, but the redox state of the Cu to target is not defined. The issue is that the redox state of Cu in the synaptic cleft is not known. Therefore, using a Cu(II) or Cu(I) chelator should be useless, meaning that a Cu(I) chelator is not able to remove Cu(II) from Aβ and vice versa. The first two questions were: - Which redox state of Cu ions should be targeted by the ligand to efficiently remove Cu ions from Aβ? Cu(I), Cu(II) or both? As the redox state of Cu is not known, the ligand should target both of them. Thus, it would be interesting to compare the in vivo efficiency of a Cu(II), a Cu(I) and a Cu(I/II) chelator. - Is it possible to target both redox states? If yes, is there a risk that the Cu-complex formed with the ligand itself can produce efficiently ROS by cycling between Cu(I) and Cu(II)? What parameters should be considered? It can be possible to chelate both redox states and there is a risk to produce ROS, but this is not mandatory. Note that chelating both redox states could also be done by a mixture of a Cu(II) and a Cu(I) chelator. In this part, a ligand L has been studied. It is able to chelate both Cu(II) and Cu(I) and there is no more production of ROS when L is added to Cu-Aβ. The hypothesis to explain this

~ 179 ~

General conclusion phenomenon is the difference between the Cu(I) coordination and geometry and the Cu(II) ones. Indeed, the coordinating ligands are not exactly the same as well as the geometry. It takes time to switch from one Cu complex (e.g. Cu(I) complex) to the other one (e.g. Cu(II) complex). This can explain the fact that the ligand can chelate both redox states without producing ROS. This shed light also the importance of the design of the ligand.

The last part of this thesis focuses on the impact of Zn ions on the Cu chelation since the concentration of Zn ions in the synaptic cleft is much higher than the one of Cu ions. The third proof of concept focuses on the thermodynamics (Figure 1, orange and purple). In the absence of Zn(II), a higher affinity constant for Cu of the ligand than Aβ is sufficient to remove Cu ions from Aβ (if the kinetic is favourable). - Is it also the case in the presence of Zn ions? Based on the results obtained with the ligands L2 and Lc, Zn(II) can preclude the Cu(II) removal from Aβ while in the absence of Zn(II), the Cu(II) removal by the ligand was possible. Both L2 and Lc have higher affinity constants for Cu and Zn than the ones of Aβ. The difference between themselves is the selectivity of Cu(II) over Zn ions (with the selectivity being the ratio between the affinity constant for Cu(II) and the one for Zn(II)). The selectivity for L2 is higher than the one for Aβ while the selectivity for Lc is lower than the one for Aβ. Both L2 and Lc can remove Cu(II) from Aβ, while in the presence of Zn ions, only L2 is able. This part highlights the importance of the selectivity of Cu over Zn ions on the Cu chelation. Not only the affinity constant is important, but also the selectivity compared to one for Aβ.

Finally, the last proof of concept describes the “pull-push” effect (Figure 1, red). The pull-push effect is the fact that for a category of ligand, Zn ions pull Cu ions out from the Aβ and push them inside the ligand. - Is a ligand with an affinity constant for Cu in the same range than Aβ able to remove totally Cu from Aβ? Can Zn(II) help the ligand to chelate Cu ions from Aβ in this case? The “pull-push” concept illustrates that a Cu chelator, with an affinity constant around the one of Aβ, can remove almost all Cu(II) bound to Aβ, only in the presence of Zn(II). Indeed, the ligands for the “pull-push” need an affinity constant for Cu in the same range than the one for Aβ and a selectivity of Cu over Zn ions higher than the one for Aβ. Without Zn(II), Cu ions are bound to Aβ and to the ligand in an equilibrium. In the presence of Zn(II), when the “pull-push” effect occurs, more Cu ions are bound to L than to Aβ, for the same ligand. This can be useful in the AD context: le ligand is not able to remove « healthy » Cu ions except in a Zn- rich environment such as the synaptic cleft, where Cu ions are toxic.

~ 180 ~

General conclusion

Then, it would be interesting to investigate the impact of the selectivity regarding Cu(I) on the Cu removal from Aβ. Indeed, as Zn ions impact the Cu(II) chelation by a ligand, they could also impact the chelation of Cu(I). Furthermore, the impact of Zn(II) on the kinetic of Cu binding is also an important parameter to study. Preliminary experiments have been performed and show that in vitro, Zn can preclude the Cu chelation. Indeed, Zn ions can be first chelated by the ligand and the dissociation time is very long, meaning that Cu will stay bound to Aβ. Other studies on the kinetic issue are the addition of arms on chelators such as ABH. Indeed, with this kind of ligands, Cu(II) needs to deprotonate an amide to be chelated. This takes time and the addition of chelating arms can be of help in the Cu removal from Aβ. Regarding the “pull-push” concept, the study of the second approach, i.e. the ligand with a very weak affinity constant for Cu(II) and a quite good selectivity, would be interesting as well as the study of the “pull-push” Cu(I)-Zn. Another investigation regarding the Cu chelation is the use of a Mn-complex as a pro-chelator. The idea is to use a superoxide dismutase mimic complex able to stop the ROS production as a ligand against AD. Preliminary experiments show that a swap of metallic ions, Cu and Mn, between Aβ and the complex occurs, stopping the associated deleterious events of Cu-Aβ. Furthermore, the metallophoric capabilities of the ligand should also be studied in order to re- equilibrate the homeostasis in the brain.

Later, the idea would be to gather these criteria in only one ligand with the already known criteria (BBB permeability, the intrinsic toxicities of the ligand and the Cu-complex): a fast Cu(I) and Cu(II) chelation, quite low affinity constants for Cu(I) and Cu(II) in order to not remove “healthy” Cu(II) and Cu(I), a Cu over Zn selectivity allowing the Cu(I) and Cu(II) chelation only in the presence of Zn ions. Then, it will be interesting to understand if other criteria are needed or not. Actually, the answer would be yes, it is needed more criteria for an efficient ligand in vivo. For example, the Cu(II)-complex BBB permeability, its metabolism and finally its excretion have to be studied. Furthermore, as Zn ions have an important impact on the Cu chelation, other biomolecules or cations can also alter it. It would interesting to study the impact of Ca(II), Fe(II/III) for example or also some amino acids, the acetylcholine, etc., on the Cu chelation from thermodynamic and kinetic aspects.

~ 181 ~

Annexes

Annexes

A- Determination of the affinity constant of Aβ peptide for Cu(II)

This section focused on the determination of the affinity constant of Aβ peptide for Cu(II), using a new water soluble UV-Visible dye. It is composed of a summary of the article published in Analytical Chemistry in 2017, the publication itself and finally the supporting information. The results reported have been obtained in collaboration with two groups. The organic and inorganic synthesises as well as the crystallographic characterisation of the water-soluble dye have been performed by Laurent Lisnard and his colleagues (at the University Pierre et Marie Curie, in Paris), and the fitting of the EXAFS experiment by Stéphanie Sayen and Emmanuel Guillon (at Institut de chimie moléculaire, in Reims). I have co-written the first draft of the paper with Valentina Borghesani.

A-i. Summary

This article reports the use of a new water soluble UV-Visible dye L (Figure 1) to determine the affinity constant of Aβ peptide for Cu(II). This is an important parameter for the design of ligands in a therapeutic approach relying on chelation because the chelators need an affinity constant for Cu(II) higher than Aβ. Many works were previously reported and the affinity constant values obtained ranged from 106 M-1 to 1019 M-1 at pH 7.4. More recently, a consensual have been proposed by our group around 1010 M-1 at pH 7.4. Most of these studies used potentiometric titrations, isothermal calorimetry, and fluorescence or competition experiments. In the present article we describe the use of a competition experiment with a new UV-Visible dye that confirms the previous value. In addition, the affinity constants for Cu(II) of mutated Aβ peptides are determined with this method; the results obtained are consistent with the coordination site of Cu(II) to Aβ at pH 7.1.

The first part of the study focuses on the organic and inorganic synthesises of the UV-Visible competitor L (Figure 1), with a crystallographic characterisation of the Cu(II) complex. EXAFS and EPR experiments are performed to define the Cu(II) coordination in solution. Both in crystals and in solution, this complex has the same coordination mode: two nitrogens and two oxygens bind the metal ion. The Cu(II)-L complex exhibit an intense absorbance band at 330 nm.

~ 182 ~

Annexes

Figure 1. Scheme of the ligand L.

A second part is dedicated to the determination of the affinity constant of the UV-Visible dye L for Cu(II), at pH 7.1. Two three-amino-acid peptides, ABH and BAH (B stands for β-alanine) which affinity constant for Cu(II) are known, are used for a competition experiment. L is first added in the solution, then Cu(II), and after around 30 minutes, one of the peptides is added. When the thermodynamic equilibrium is reached, another equivalent of the peptide is added, and this until 10 equivalents of peptides. During the experiment, the absorbance of the Cu(II)-L complex is followed, and plotted as a function of the number of added equivalent of peptides. Then, with an home-made fitting, the affinity constant of L for Cu(II) is determined and the average value is 3.2 ± 1.0 x 109 M-1 at pH 7.1.

A third part of this article focuses on the main issue of this work: the determination of the affinity constant of Aβ for Cu(II). Because the first sixteen amino-acids are responsible of the metal ions chelation, the competition experiment is performed with the Aβ16. The same competition, changing the ABH and BAH peptides by the Aβ peptide, and the same home-made fitting than for the determination of the affinity constant for Cu(II) of L are performed. The average value is 1.6 x 109 M-1. This value is in line with previous report proposed by Kowalik-Jankowska et al. where potentiometric titration was used. Then, the affinity constants of Aβ28 and Aβ40 have been evaluated. As they are closed to the one of Aβ16, Aβ16 is a good model of the entire peptide with respect to Cu(II) coordination.

The last part is about the coordination of Cu(II) with Aβ. A wide series of Aβ16 mutants is used to determine which amino acid is involved in the Cu(II) coordination (if the affinity constant of the mutant changes, this is because the amino acid mutated is involved in the coordination). Note that the Cu(II) coordination is already well known. Therefore, if with this technique, the same Cu(II) coordination is proposed, this competition method with L can be claimed as a robust method. A first kind of mutant is studied: the N-terminal amine involved in the Cu(II) coordination is acetylated, preventing its Cu(II) binding ability. The impact of this mutation on the affinity constant is very important and the value cannot be determined precisely by the competition with L due to its too low value. This impact confirms that the N-terminal amine is strongly involved in the Cu(II) coordination. A second class of mutants is the carboxylate to amide change, carboxylate residues coordinating the metal ion in an

~ 183 ~

Annexes apical position. The affinity constants obtained for these mutants are lower than those of the wild- type Aβ, but higher than the acetylated peptide. This result is consistent with the literature: the apical position is less important in the Cu(II) coordination than the equatorial position. Furthermore, for the apical position, 4 ligands are in exchange, thus one mutation of a carboxylate group affects the Cu(II) coordination but not in an important way because three other ones remain. A last type of mutations is the Histidine to Alanine or to Arginine mutations. Cu(II) is coordinated by His6 and His13 or His14, in a dynamic exchange. The affinity constants measured with the competition experiment are in line with this coordination binding: the values for the peptides with the mutation H6A or H6R are close to each other and lower than those of the peptide with H13A and H14A. His6 has a stronger impact on the Cu(II) coordination with Aβ. Another study is performed with the murine Aβ. This peptide has 3 mutations compared to the human one: R5G, Y10F and H13R. The affinity constant for Cu(II) of the human mutants of Aβ corresponding to the mutations involved in the murine peptide are also evaluated with this method. Consistent results are also obtained. Indeed, the affinity constant of the murine peptide is twice stronger than the one of the human Aβ, as proposed in the literature. The R5G mutation is the most important one and the affinity constant of the associated mutant is the same than the murine. This is due to the formation of the 6-membered metallacycle between the peptidic bond Gly5-His6 and the His6. This formation of metallacyle stabilizes the Cu(II) complex, increasing its affinity constant.

These experiments allow to propose that L is a good competitor to determine the affinity constant of peptides or proteins with a moderate affinity constant for Cu(II). The validation of the Cu(II) coordination using different mutants also supports the reliability of L. This method is easy to handle do to the important absorbance of the Cu(II)-L complex. Indeed, only a UV-Vis spectrophotometer is required and a small quantity of L and peptide is needed. Furthermore, by UV-Vis, there is no bleaching of the dye, no inner filter effect, etc. Note that for the determination of the affinity constant of a peptide at another pH than pH 7.1, the affinity constant for the competitor L has to be determined previously. For peptides or proteins with lower or higher affinity constants, modifications of L are needed to adapt its affinity constant in the appropriate range.

This article proposes a consistent affinity constant of Aβ for Cu(II), at 1.6 x 109 M-1. This is an important parameter to know, for the development of new ligands able to remove Cu(II) from Aβ in a therapeutic approach based on chelation.

~ 184 ~

Annexes

A-ii. Article

~ 185 ~

Annexes

~ 186 ~

Annexes

~ 187 ~

Annexes

~ 188 ~

Annexes

~ 189 ~

Annexes

~ 190 ~

Annexes

~ 191 ~

Annexes

~ 192 ~

Annexes

A-iii. Supporting information

~ 193 ~

Annexes

~ 194 ~

Annexes

~ 195 ~

Annexes

~ 196 ~

Annexes

~ 197 ~

Annexes

~ 198 ~

Annexes

B- Zn(II) coordination to Aβ peptide

This section focuses on the Zn(II) coordination to the Aβ peptide. It is composed of a summary of the article published in Inorganic Chemistry in 2016, the publication itself and finally the supporting information. My contribution to this work has been some NMR experiments and the analysis of the NMR results and participation in the writing.

B-i. Summary

This article describes the Zn(II) coordination to the Aβ peptide at pH 7.4. The study of the interaction between Zn(II) and Aβ peptide is biologically relevant since Zn(II) is the most common transition metal ion involved in the neurotransmission signal. It is present in higher concentrations in the synaptic cleft than Cu. In addition, important concentrations of Zn(II) have been found in the senile plaques, evidencing the interactions between Zn(II) and Aβ. Previous works have described a possible protective role of Zn(II) ion, such as the precipitation of Aβ in excess into redox-silent species. Moreover, as previously explained, Zn(II) has also an impact on the Cu(II) coordination, on the ROS production by the Cu-Aβ complex and on the aggregation of the peptide; it is thus important to understand its interaction with the peptide in order to understand its impact in the disease. Zn(II) coordination models already exist, but are still debated. Investigation of Zn(II) coordination is a difficult task since Zn(II) is silent in most of the spectroscopic technics and the flexibility of Aβ peptide precludes X-Ray characterization. In this work, XAS and NMR studies are paralleled; a first EXAFS study determines the number of ligands coordinating the metal ion. Then NMR and XANES studies with a wide series of mutated Aβ peptides allow the characterization of the amino acids involved in Zn(II) binding. The affinity constant is also discussed.

The first part of this work focuses on the characterisation of the Zn(II) binding site to Aβ. An EXAFS study allows to determine the number and the nature of the atoms involved in the metal ion coordination and also the distance between the metal centre and the coordinating atoms. However, due to the ill-defined coordination sphere of Zn(II), only the first shell can be solved. The fitting of the experimental data proposes a 4N/O shell at an average 1.98 Å distance from Zn(II). A complementary NMR study investigates also the binding site of Zn(II), using comparisons between the spectrum of Aβ and the one of the Zn(II)-Aβ complex. Two kind of modifications of the spectra are observed due to the addition of Zn(II) to Aβ: the broadening of some protons and their up- and down-field shift. Protons in the close vicinity of the atom bound to Zn(II) due to Lewis acidity of Zn ion have their signal modified. In addition, remote protons can also be impacted since chelation of Zn(II) by Aβ leads to modifications

~ 199 ~

Annexes in the “folding” of the peptide, and the protons, even not involved in the coordination, exhibit different signals than the ones for Aβ. Hence, signals associated to more than four different amino-acid residues, are affected by the presence of Zn(II) and so the coordination site cannot be straightforwardly characterized.

The second part of this article presents the XANES and NMR studies of a wide series of Aβ modified peptides. For the XANES experiment, comparison of the Zn(II)-Aβ spectrum and of the Zn(II)-mutant spectrum allows to identify the amino acids involved in the coordination of the metal ion. First of all, the white line intensity of the XANES data is in good agreement with the four-coordination of Zn(II). Furthermore, the data are consistent with biological systems, with a four-coordination including two Histidine, except for the Zn(II)-E11Q complex. If one Histidine is mutated, it is replaced by another one; two Histidine being always involved in the Zn(II) coordination. For the E11Q mutant, the carboxylate group is replaced by a Histidine, leading to a four-coordination with three Histidine. In summary, the XANES data describe a four-coordination system, involving two Histidine for the native peptide. The main amino acids inducing changes in the XANES fingerprints are the mutations of the Histidine (H6A, H13A and H14A) and the mutation E11Q. Some weak changes upon the Zn(II) addition are induced by the N-terminal acetylation and by the D1N, E3Q and D7N mutations.

NMR study is needed to gain more insights in the coordination of Zn(II) to Aβ. The NMR spectra are compared between each other and there are two possibilities: (i) if the changes induced by Zn(II) on Aβ fingerprint and on the mutant one are the same, the amino acid mutated is not involved in the Zn(II) coordination, (ii) if these changes are different, the amino acid is important for the coordination. The first important point of this study is the determination of which Histidine is involved in the binding of the metal ion. H13A and H14A have shown the broadening of R5 (mutants in which H6 can coordinate the metal ion), which disappeares for the H6A, evidencing the importance of H6 in the coordination. Moreover, the affinity constant of H6A for Zn(II) is lower than H13A and H14A. Regarding H13, there is no important broadening upon Zn(II) addition for the mutation H13A while for the wild type Aβ there is. H13 should be involved in the Zn(II) coordination. In addition, V12 signal is affected for H6A and H14A (mutants in which H13 can coordinate the metal ion) but is not for H13A. This evidences the implication of H13 in the Zn(II) coordination. All of these data lead to the conclusion that H6 is involved in the Zn(II) coordination, but H13 and H14 are in a dynamic exchange (Figure 2). The other important point of this part is the determination of the importance of the carboxylate groups. The carboxylate groups of Asp1, Glu3 and Asp7 are in a dynamic exchange for the binding of Zn(II), whereas the carboxylate of Glu11 binds Zn(II) independently (Figure 2). The last important point of this part is the importance of the N-terminal amine. The changes induced by the addition of Zn(II) for the

~ 200 ~

Annexes

Aβ peptide and for the N-terminal acetylated are identical, meaning that the changes are due to the involvement of the carboxylate side chain the Asp1. This is in line with the weak decrease of the affinity constant for Zn(II) of the N-terminal acetylated peptide. Contrary to what is previously proposed in the literature, the N-terminal amine is not involved in the Zn(II) coordination at pH 7.4 (Figure 2), but seems to be involved at higher pH (around 9).

Figure 2. Scheme of the Zn(II) coordination site to Aβ16.

In conclusion, this article describes the Zn(II) coordination mode to Aβ at pH 7.4. His6 and Glu11 bind Zn(II). The carboxylate group of Asp1, Glu3 and Asp7 are in a dynamic exchange for one binding position, as well as the His13 and the His14 for another one. This binding site is different from the Cu(II) one. This is in line with a largely stronger affinity for Cu(II) compared to Zn(II) (4 order of magnitude) This impacts also the global charge of these complexes, since Cu(II) is bound to the deprotonated N- terminal amine whereas Zn(II) is not bound and the amine is protonated. This difference could participate in the distinct behaviour of Cu and Zn-induced Aβ aggregation.

~ 201 ~

Annexes

B-ii. Article

~ 202 ~

Annexes

~ 203 ~

Annexes

~ 204 ~

Annexes

~ 205 ~

Annexes

~ 206 ~

Annexes

~ 207 ~

Annexes

~ 208 ~

Annexes

~ 209 ~

Annexes

~ 210 ~

Annexes

~ 211 ~

Annexes

~ 212 ~

Annexes

B-iii. Supporting information

~ 213 ~

Annexes

~ 214 ~

Annexes

~ 215 ~

Annexes

~ 216 ~

Annexes

~ 217 ~

Annexes

~ 218 ~

Annexes

~ 219 ~

Annexes

~ 220 ~

Annexes

~ 221 ~

Annexes

~ 222 ~

Annexes

~ 223 ~

Annexes

~ 224 ~

Annexes

~ 225 ~

Annexes

~ 226 ~

Annexes

C- The first Cu(I) chelator against AD

This section is dedicated to the description of the first Cu(I) chelator reported within the context of Alzheimer’s disease. It is composed of a summary of the article published in Metallomics in 2015, the publication itself and finally the supporting information. This work has mainly been performed by Elena Atrián-Blasco and her collegues in Spain. My contribution is the aggregation part and in the visualisation of the samples by AFM.

C-i. Summary

This article describes for the first time the use of a Cu(I) chelator in order to remove the metal ion from the Aβ peptide and to reduce the associated deleterious effects. As demonstrated in the section II-A, most of the researches on metal chelation or metal redistribution approaches focus on the Cu(II) ion. Nevertheless, until now, there is no evidence of the redox state of Cu ions in the synaptic cleft. In addition, some studies have proved that a Cu(II) ligand naturally present in the brain, Human Serum Albumin (HSA), protected less cells from Cu-Aβ complex toxicity in comparison with metallothionein 3 (MT-3), which is a Cu(I) ligand.

This work proposes a Cu(I) chelator, the Phosphane 1,3,5-Triaza-7-phosphaAdamantane (PTA), which is a phosphine based ligand strongly resistant to oxidation. Another advantage is its water solubility, usefull in the Alzheimer’s disease studies. It is also biocompatible and has a low intrinsic toxicity.

The first step of this work is the removal of Cu ions from the Aβ peptide. For this study, XANES and EPR experiments were performed. XANES allows to follow Cu(II) and Cu(I) at the same time, and EPR is specific to Cu(II) ions, so in the presence of Cu(I), there is no signal. XANES experiments are used to monitor the removal of Cu(I) from Aβ by PTA. The ligand was added progressively and after 4.5 equiv. of PTA, Cu ion is totally chelated by the ligand and not by the peptide. 4 equiv. are needed for the Cu(I) chelation and 0.5 equiv. is required for the reduction of Cu(II) into Cu(I). EPR experiment confirms this result: with 6 equiv. of PTA, the signature of Cu(II)-Aβ becomes flat, in line with complete Cu(II) reduction. The results of the UV-Visible competition are consistent with that: the addition of PTA eliminates the d-d band of Cu(II)-Aβ; it is needed around 1 h to remove Cu(I) from Aβ. NMR experiments are in good agreement with these results of Cu ion removal. Indeed, upon Cu(I) addition, the spectrum of Aβ changes mainly for the Histidine signals. When PTA is added to Cu(I)-Aβ, the signal for apo Histidine are recovered. 4.5 equiv. of PTA are needed to remove Cu ion from the Aβ peptide,

~ 227 ~

Annexes reduce it and chelate it. Another aspect of this removal is the oxidation of PTA, but oxidised PTA cannot chelate the metal ion, that is why 4.5 equiv. of PTA are needed.

The second part of the work is the analysis of the impact of PTA on the ROS production. The following of the ascorbate consumption kinetic as well as the detection of the HO• formed by the kinetic of fluorescence of 7-OH-CCA are used mirroring the ROS production. Experiments are performed with and without Aβ. For the Asc consumption experiments, for both cases, the first part of the curves shows an important decrease of their gradient upon addition of PTA, meaning that PTA is able to decrease if not stop the ROS production by Cu-Aβ complexes. Note that PTA has not a high enough affinity constant for Cu(I) to fully avoid free Cu. By fluorescence, for the experiments without Aβ, the more equiv. of PTA is added, the lower is the fluorescence at the plateau. This means that PTA has captured HO• radicals instead of the 3-CCA, leading to a lower concentration of 7-OH-CCA at the end of the reaction. Thus, the phosphine group of PTA is oxidised into phosphoryl group, the phosphorous is not yet able to coordinate Cu(I) and there is a release of the metal ions. Free Cu ions

• can react with ascorbate and O2 to trigger the HO . This hypothesis is relevant due to the exactly same gradient of the fluorescence curve of Cu alone and of Cu in the presence of PTA. This PTA oxidative process is evidence by the increase of the lag phase: when the concentration of PTA increases, there are more PTA available for the Cu(I) chelation, and so, when PTA is oxidised, another PTA can chelate Cu(I) and increase the protection against the ROS production. In the presence of Aβ peptide, the same trend is observed: PTA induced less 7-OH-CCA formation and there is a delay due to oxidation of PTA itself. In conclusion, PTA is able to stop the ROS production but it is required a huge quantity of ligand to not forming ROS after a while. These results have also been confirmed by the ascorbate consumption experiments.

The last part of this work focuses on the aggregation of the Aβ peptide and the impact of PTA on it. ThT fluorescence experiments were performed and the samples were imaged by AFM. Apo Aβ peptide aggregates into fibrils whereas Cu(II)-Aβ aggregates into oligomers, non-fibrillar species. In the presence of PTA and Cu(II), the aggregation of Aβ looks like the apo aggregation: PTA sequesters the metal ions and the Aβ peptide can aggregate into fibrils. Nevertheless, 4.5 equiv. of PTA is not enough here. Indeed, Cu(I)-PTA4 complex is not air-stable for a long time and can be oxidised by O2. A release of Cu ions thus happens and a Cu(II) induced Aβ aggregation can occur. It is very important to use a super-stoichiometric ratio of Cu:PTA to keep the Cu(I)-PTA4 during all the aggregation process. For this experiment, 20 equiv. of PTA are added. Thus, PTA is able to prevent the Cu(II)-Aβ oligomeric species, supposed to be the most toxic forms.

~ 228 ~

Annexes

In conclusion, this article describes the first Cu(I) chelator able to remove Cu(II) and Cu(I) from Aβ, to reduce considerably the ROS production by Cu-Aβ complex (see Scheme below) and to avoid the formation of the toxic oligomeric species in a large excess of ligand. Moreover, as PTA can be easily modified by alkylation or arylation in the nitrogen atoms, it could be interesting to improve the PTA scaffold to be more resistant to oxidation, more air-stable, and to increase its affinity constant for Cu(I), making dimer of PTA for instance, in order to remove all the Cu ions from the Aβ peptide. Another improvement could be the coupling of PTA and an Aβ target moiety, in order to be more specific in the removal of Cu ions.

Scheme representing the strategy used in this publication: a Cu(I) chelator is used on order to remove Cu ions from Aβ and to stop the ROS production.

~ 229 ~

Annexes

C-ii. Article

~ 230 ~

Annexes

~ 231 ~

Annexes

~ 232 ~

Annexes

~ 233 ~

Annexes

~ 234 ~

Annexes

C-iii. Supporting information

~ 235 ~

Annexes

~ 236 ~

Annexes

~ 237 ~

Annexes

~ 238 ~

Annexes

~ 239 ~

Annexes

~ 240 ~

Annexes

~ 241 ~

Annexes

~ 242 ~

Annexes

~ 243 ~

Annexes

~ 244 ~

Annexes

~ 245 ~

Annexes

~ 246 ~

Annexes

~ 247 ~

Annexes

~ 248 ~

Annexes

~ 249 ~

Annexes

~ 250 ~

Résumé

La maladie d’Alzheimer est une maladie neurodégénérative, touchant plus de 30 millions de personnes dans le monde. A ce jour, seules des thérapies symptomatiques sont disponibles ; aucun traitement curatif n’existe. Une des hypothèses concernant cette maladie propose une mauvaise régulation des quantités en ions métalliques, notamment les ions Cu et Zn, dans certaines zones du cerveau. Ils favoriseraient une accumulation de peptides appelés Amyloïdes-β (Aβ) dans les fentes synaptiques. Ces dépôts empêcheraient les connexions neuronales, entrainant les symptômes connus de la maladie, tels que la perte de mémoire ou les déficiences intellectuelles. Les ions Cu seraient également responsables d’un stress oxydant incontrôlé, dégradant entre autres les membranes neuronales. Les ions Cu sont donc une cible thérapeutique à privilégier. Les recherches se dirigent vers le développement de nouvelles molécules, dites chélateurs, en vue d’extraire sélectivement ces ions Cu (par rapport aux ions Zn), pour réguler leur quantité et limiter voire empêcher cette accumulation de peptides. Mon projet de recherche se place précisément dans ce contexte. Différents chélateurs des ions Cu(II) et Cu(I) sont étudiés, en présence ou non de Zn(II), pour comprendre les paramètres à prendre en compte pour le développement de chélateurs efficaces. La première partie de cette étude regroupe différentes preuves de concept concernant les chélateurs des ions Cu. L’aspect cinétique du retrait du Cu(II) du peptide Aβ par un chélateur est étudié grâce à des ligands macrocycliques. Ensuite, l’état d’oxydation des ions Cu dans les fentes synaptique n’étant pas connu à ce jour, deux chélateurs du Cu(I) ou du Cu(I/II) sont proposés. La seconde partie de l’étude prend en compte l’impact du Zn(II) dans la chélation des ions Cu. Le côté thermodynamique de la chélation du Cu en présence de Zn(II) est mis en évidence grâce à différents chélateurs aux caractéristiques différentes.

Abstract

Alzheimer’s disease is a neurodegenerative disease, affecting more than 30 million people all over the world. Nowadays, only symptomatic therapies exist, there is no cure yet. A dyshomeostasis of metal ions such as Cu and Zn ions in some areas of the brain is one of the different hypothesis about this disease. They would promote an accumulation of peptides, the Amyloid-β (Aβ) peptides, in the synaptic cleft. These aggregates would prevent the neuronal connections, triggering known symptoms of the disease, such as memory loss or cognitive impairments. Cu ions would also be responsible for an important oxidative stress, destroying the neuronal membranes for example. Cu ions are an important therapeutic target to cure the disease. Investigations are currently focusing on the development of new molecules, called chelators, in order to remove selectively Cu ions (over Zn ions), to regulate their concentrations and avoid the accumulation of the peptides. My research project focuses precisely on such kind of investigations. Different Cu(II) and Cu(I) chelators are studied, in the presence or not of Zn(II), in order to understand the different criteria to take into account for the development of good chelators. Different proof-of-concepts are developed in the first part. The kinetic aspect of the removal of Cu(II) from the Aβ peptide by a chelator is studied with macrocyclic ligands Then, the redox state of Cu ions in the synaptic cleft staying unknown, two Cu(I) or Cu(I/II) chelators are proposed. The second part of the study takes into account the impact of Zn(II) in the Cu chelation. The thermodynamic part of the Cu(II) chelation in the presence of Zn(II) is evidenced with different chelators.